diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml
index f3b13f7..193484e 100644
--- a/.pre-commit-config.yaml
+++ b/.pre-commit-config.yaml
@@ -10,3 +10,8 @@ repos:
         # pre-commit's default_language_version, see
         # https://pre-commit.com/#top_level-default_language_version
         language_version: python3.9
+  - repo: https://github.com/pycqa/isort
+    rev: 5.12.0
+    hooks:
+      - id: isort
+        args: ["--profile", "black", "--filter-files"]
\ No newline at end of file
diff --git a/pyproject.toml b/pyproject.toml
index a852e18..2645631 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -44,7 +44,13 @@ test = [
 dev = [
     "bandit",
     "black==22.12.0",
+    "isort",
     "pre-commit",
     "pylint",
     "pycodestyle"
 ]
+
+[tool.isort]
+profile = "black"
+src_paths = ["src", "tests"]
+skip_glob = [".aws_sam"]
\ No newline at end of file
diff --git a/scripts/credential_management.py b/scripts/credential_management.py
index 6daba3a..5b3b6f6 100755
--- a/scripts/credential_management.py
+++ b/scripts/credential_management.py
@@ -6,7 +6,6 @@
 import sys
 
 import boto3
-
 from requests.auth import _basic_auth_str
 
 
diff --git a/scripts/cumulus_upload_data.py b/scripts/cumulus_upload_data.py
index 090bba0..6975c80 100755
--- a/scripts/cumulus_upload_data.py
+++ b/scripts/cumulus_upload_data.py
@@ -4,7 +4,6 @@
 import argparse
 import os
 import sys
-
 from pathlib import Path
 
 import boto3
diff --git a/scripts/migrate_versioning.py b/scripts/migrate_versioning.py
new file mode 100644
index 0000000..3b555d6
--- /dev/null
+++ b/scripts/migrate_versioning.py
@@ -0,0 +1,90 @@
+""" Utility for adding versioning to an existing aggregator data store 
+
+This is a one time thing for us, so the CLI/Boto creds are not robust.
+"""
+import argparse
+import boto3
+import io
+import json
+
+from rich import progress
+
+UPLOAD_ROOT_BUCKETS = ["archive", "error", "last_valid", "latest", "site_upload"]
+
+
+def _get_s3_data(key: str, bucket_name: str, client) -> dict:
+    """Convenience class for retrieving a dict from S3"""
+    try:
+        bytes_buffer = io.BytesIO()
+        client.download_fileobj(Bucket=bucket_name, Key=key, Fileobj=bytes_buffer)
+        return json.loads(bytes_buffer.getvalue().decode())
+    except Exception:  # pylint: disable=broad-except
+        return {}
+
+
+def _put_s3_data(key: str, bucket_name: str, client, data: dict) -> None:
+    """Convenience class for writing a dict to S3"""
+    b_data = io.BytesIO(json.dumps(data).encode())
+    client.upload_fileobj(Bucket=bucket_name, Key=key, Fileobj=b_data)
+
+
+def _get_depth(d):
+    if isinstance(d, dict):
+        return 1 + (max(map(_get_depth, d.values())) if d else 0)
+    return 0
+
+
+def migrate_bucket_versioning(bucket: str):
+    client = boto3.client("s3")
+    res = client.list_objects_v2(Bucket=bucket)
+    contents = res["Contents"]
+    moved_files = 0
+    for s3_file in contents:
+        if s3_file["Key"].split("/")[0] in UPLOAD_ROOT_BUCKETS:
+            key = s3_file["Key"]
+            key_array = key.split("/")
+            if len(key_array) == 5:
+                key_array.insert(4, "000")
+                new_key = "/".join(key_array)
+                client.copy({"Bucket": bucket, "Key": key}, bucket, new_key)
+                client.delete_object(Bucket=bucket, Key=key)
+                moved_files += 1
+    print(f"Moved {moved_files} uploads")
+
+    study_periods = _get_s3_data("metadata/study_periods.json", bucket, client)
+
+    if _get_depth(study_periods) == 3:
+        new_sp = {}
+        for site in study_periods:
+            new_sp[site] = {}
+            for study in study_periods[site]:
+                new_sp[site][study] = {}
+                new_sp[site][study]["000"] = study_periods[site][study]
+        _put_s3_data("metadata/study_periods.json", bucket, client, new_sp)
+        print("study_periods.json updated")
+    else:
+        print("study_periods.json does not need update")
+
+    transactions = _get_s3_data("metadata/transactions.json", bucket, client)
+    if _get_depth(transactions) == 4:
+        new_t = {}
+        for site in transactions:
+            new_t[site] = {}
+            for study in transactions[site]:
+                new_t[site][study] = {}
+                for dp in transactions[site][study]:
+                    new_t[site][study][dp] = {}
+                    new_t[site][study][dp]["000"] = transactions[site][study][dp]
+        _put_s3_data("metadata/transactions.json", bucket, client, new_t)
+        print("transactions.json updated")
+    else:
+        print("transactions.json does not need update")
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser(
+        description="""Util for migrating aggregator data"""
+    )
+    parser.add_argument("-b", "--bucket", help="bucket name")
+    args = parser.parse_args()
+    migrate_bucket_versioning(args.bucket)
diff --git a/src/handlers/dashboard/get_chart_data.py b/src/handlers/dashboard/get_chart_data.py
index 370a86e..3150965 100644
--- a/src/handlers/dashboard/get_chart_data.py
+++ b/src/handlers/dashboard/get_chart_data.py
@@ -2,8 +2,7 @@
 This is intended to provide an implementation of the logic described in docs/api.md
 """
 import os
-
-from typing import List, Dict
+from typing import Dict, List
 
 import awswrangler
 import boto3
diff --git a/src/handlers/shared/functions.py b/src/handlers/shared/functions.py
index 45774fe..8cd2462 100644
--- a/src/handlers/shared/functions.py
+++ b/src/handlers/shared/functions.py
@@ -1,10 +1,9 @@
 """ Functions used across different lambdas"""
 import io
-import logging
 import json
-
-from typing import Dict, Optional
+import logging
 from datetime import datetime, timezone
+from typing import Dict, Optional
 
 import boto3
 
@@ -75,6 +74,7 @@ def update_metadata(
     site: str,
     study: str,
     data_package: str,
+    version: str,
     target: str,
     dt: Optional[datetime] = None,
     meta_type: str = JsonFilename.TRANSACTIONS.value,
@@ -84,18 +84,20 @@ def update_metadata(
     if meta_type == JsonFilename.TRANSACTIONS.value:
         site_metadata = metadata.setdefault(site, {})
         study_metadata = site_metadata.setdefault(study, {})
-        data_package_metadata = study_metadata.setdefault(
-            data_package, TRANSACTION_METADATA_TEMPLATE
+        data_package_metadata = study_metadata.setdefault(data_package, {})
+        version_metadata = data_package_metadata.setdefault(
+            version, TRANSACTION_METADATA_TEMPLATE
         )
         dt = dt or datetime.now(timezone.utc)
-        data_package_metadata[target] = dt.isoformat()
+        version_metadata[target] = dt.isoformat()
     elif meta_type == JsonFilename.STUDY_PERIODS.value:
         site_metadata = metadata.setdefault(site, {})
-        study_period_metadata = site_metadata.setdefault(
-            study, STUDY_PERIOD_METADATA_TEMPLATE
+        study_period_metadata = site_metadata.setdefault(study, {})
+        version_metadata = study_period_metadata.setdefault(
+            version, STUDY_PERIOD_METADATA_TEMPLATE
         )
         dt = dt or datetime.now(timezone.utc)
-        study_period_metadata[target] = dt.isoformat()
+        version_metadata[target] = dt.isoformat()
     return metadata
 
 
diff --git a/src/handlers/site_upload/fetch_upload_url.py b/src/handlers/site_upload/fetch_upload_url.py
index 41d5fc9..f1ef8e8 100644
--- a/src/handlers/site_upload/fetch_upload_url.py
+++ b/src/handlers/site_upload/fetch_upload_url.py
@@ -46,6 +46,6 @@ def upload_url_handler(event, context):
     res = create_presigned_post(
         os.environ.get("BUCKET_NAME"),
         f"{BucketPath.UPLOAD.value}/{body['study']}/{body['data_package']}/"
-        f"{metadata_db[user]['path']}/{body['filename']}",
+        f"{body['version']}/{metadata_db[user]['path']}/{body['filename']}",
     )
     return res
diff --git a/src/handlers/site_upload/powerset_merge.py b/src/handlers/site_upload/powerset_merge.py
index 7ac8e97..d1c768a 100644
--- a/src/handlers/site_upload/powerset_merge.py
+++ b/src/handlers/site_upload/powerset_merge.py
@@ -3,19 +3,17 @@
 import logging
 import os
 import traceback
-
 from datetime import datetime, timezone
 
 import awswrangler
 import boto3
 import pandas
-
 from numpy import nan
 from pandas.core.indexes.range import RangeIndex
 
+from src.handlers.shared.awswrangler_functions import get_s3_data_package_list
 from src.handlers.shared.decorators import generic_error_handler
 from src.handlers.shared.enums import BucketPath
-from src.handlers.shared.awswrangler_functions import get_s3_data_package_list
 from src.handlers.shared.functions import (
     get_s3_site_filename_suffix,
     http_response,
@@ -47,7 +45,7 @@ def __init__(self, event):
         self.site = s3_key_array[3]
         self.study = s3_key_array[1]
         self.data_package = s3_key_array[2]
-
+        self.version = s3_key_array[4]
         self.metadata = read_metadata(self.s3_client, self.s3_bucket_name)
 
     # S3 Filesystem operations
@@ -78,7 +76,7 @@ def write_parquet(self, df: pandas.DataFrame, is_new_data_package: bool) -> None
         """writes dataframe as parquet to s3 and sends an SNS notification if new"""
         parquet_aggregate_path = (
             f"s3://{self.s3_bucket_name}/{BucketPath.AGGREGATE.value}/"
-            f"{self.study}/{self.study}__{self.data_package}/"
+            f"{self.study}/{self.study}__{self.data_package}/{self.version}/"
             f"{self.study}__{self.data_package}__aggregate.parquet"
         )
         awswrangler.s3.to_parquet(df, parquet_aggregate_path, index=False)
@@ -92,7 +90,7 @@ def write_csv(self, df: pandas.DataFrame) -> None:
         """writes dataframe as csv to s3"""
         csv_aggregate_path = (
             f"s3://{self.s3_bucket_name}/{BucketPath.CSVAGGREGATE.value}/"
-            f"{self.study}/{self.study}__{self.data_package}/"
+            f"{self.study}/{self.study}__{self.data_package}/{self.version}/"
             f"{self.study}__{self.data_package}__aggregate.csv"
         )
         df = df.apply(lambda x: x.strip() if isinstance(x, str) else x).replace(
@@ -109,7 +107,7 @@ def update_local_metadata(self, key, site=None):
         if site is None:
             site = self.site
         self.metadata = update_metadata(
-            self.metadata, site, self.study, self.data_package, key
+            self.metadata, site, self.study, self.data_package, self.version, key
         )
 
     def write_local_metadata(self):
@@ -162,9 +160,7 @@ def expand_and_concat_sets(
     site_df["site"] = get_static_string_series(None, site_df.index)
     df_copy["site"] = get_static_string_series(site_name, df_copy.index)
 
-    # TODO: we should introduce some kind of data versioning check to see if datasets
-    # are generated from the same vintage. This naive approach will cause a decent
-    # amount of data churn we'll have to manage in the interim.
+    # Did we change the schema without updating the version?
     if df.empty is False and set(site_df.columns) != set(df.columns):
         raise MergeError(
             "Uploaded data has a different schema than last aggregate",
diff --git a/src/handlers/site_upload/process_upload.py b/src/handlers/site_upload/process_upload.py
index 76b94ce..cee2b14 100644
--- a/src/handlers/site_upload/process_upload.py
+++ b/src/handlers/site_upload/process_upload.py
@@ -28,8 +28,9 @@ def process_upload(s3_client, sns_client, s3_bucket_name: str, s3_key: str) -> N
     study = path_params[1]
     data_package = path_params[2]
     site = path_params[3]
+    version = path_params[4]
     if s3_key.endswith(".parquet"):
-        if "_meta_" in s3_key:
+        if "__meta_" in s3_key:
             new_key = f"{BucketPath.STUDY_META.value}/{s3_key.split('/', 1)[-1]}"
             topic_sns_arn = os.environ.get("TOPIC_PROCESS_STUDY_META_ARN")
             sns_subject = "Process study medata upload event"
@@ -43,6 +44,7 @@ def process_upload(s3_client, sns_client, s3_bucket_name: str, s3_key: str) -> N
             site,
             study,
             data_package,
+            version,
             "last_upload",
             last_uploaded_date,
         )
@@ -56,11 +58,18 @@ def process_upload(s3_client, sns_client, s3_bucket_name: str, s3_key: str) -> N
             site,
             study,
             data_package,
+            version,
             "last_upload",
             last_uploaded_date,
         )
         metadata = update_metadata(
-            metadata, site, study, data_package, "last_error", last_uploaded_date
+            metadata,
+            site,
+            study,
+            data_package,
+            version,
+            "last_error",
+            last_uploaded_date,
         )
         write_metadata(s3_client, s3_bucket_name, metadata)
         raise UnexpectedFileTypeError
diff --git a/src/handlers/site_upload/study_period.py b/src/handlers/site_upload/study_period.py
index 276a12c..d4c73d7 100644
--- a/src/handlers/site_upload/study_period.py
+++ b/src/handlers/site_upload/study_period.py
@@ -1,15 +1,14 @@
 """ Lambda for updating date ranges associated with studies """
 
 import os
-
 from datetime import datetime, timezone
 
 import awswrangler
 import boto3
 
+from src.handlers.shared.awswrangler_functions import get_s3_study_meta_list
 from src.handlers.shared.decorators import generic_error_handler
 from src.handlers.shared.enums import BucketPath, JsonFilename
-from src.handlers.shared.awswrangler_functions import get_s3_study_meta_list
 from src.handlers.shared.functions import (
     http_response,
     read_metadata,
@@ -18,7 +17,7 @@
 )
 
 
-def update_study_period(s3_client, s3_bucket, site, study, data_package):
+def update_study_period(s3_client, s3_bucket, site, study, data_package, version):
     """gets earliest/latest date from study metadata files"""
     path = get_s3_study_meta_list(
         BucketPath.STUDY_META.value, s3_bucket, study, data_package, site
@@ -29,11 +28,13 @@ def update_study_period(s3_client, s3_bucket, site, study, data_package):
     study_meta = read_metadata(
         s3_client, s3_bucket, meta_type=JsonFilename.STUDY_PERIODS.value
     )
+
     study_meta = update_metadata(
         study_meta,
         site,
         study,
         data_package,
+        version,
         "earliest_date",
         df["min_date"][0],
         meta_type=JsonFilename.STUDY_PERIODS.value,
@@ -43,6 +44,7 @@ def update_study_period(s3_client, s3_bucket, site, study, data_package):
         site,
         study,
         data_package,
+        version,
         "latest_date",
         df["max_date"][0],
         meta_type=JsonFilename.STUDY_PERIODS.value,
@@ -52,6 +54,7 @@ def update_study_period(s3_client, s3_bucket, site, study, data_package):
         site,
         study,
         data_package,
+        version,
         "last_data_update",
         datetime.now(timezone.utc),
         meta_type=JsonFilename.STUDY_PERIODS.value,
@@ -68,11 +71,11 @@ def study_period_handler(event, context):
     s3_bucket = os.environ.get("BUCKET_NAME")
     s3_client = boto3.client("s3")
     s3_key = event["Records"][0]["Sns"]["Message"]
-    s3_key_array = s3_key.split("/")
-    site = s3_key_array[3]
-    study = s3_key_array[1]
-    data_package = s3_key_array[2]
-
-    update_study_period(s3_client, s3_bucket, site, study, data_package)
+    path_params = s3_key.split("/")
+    study = path_params[1]
+    data_package = path_params[2]
+    site = path_params[3]
+    version = path_params[4]
+    update_study_period(s3_client, s3_bucket, site, study, data_package, version)
     res = http_response(200, "Study period update successful")
     return res
diff --git a/tests/conftest.py b/tests/conftest.py
index c453c8b..09885ae 100644
--- a/tests/conftest.py
+++ b/tests/conftest.py
@@ -22,21 +22,19 @@
   aggregator
 """
 import os
-
 from unittest import mock
 
 import boto3
 import pytest
-
-from moto import mock_s3, mock_athena, mock_sns
-
+from moto import mock_athena, mock_s3, mock_sns
 from scripts.credential_management import create_auth, create_meta
+
 from src.handlers.shared.enums import BucketPath, JsonFilename
 from src.handlers.shared.functions import write_metadata
-from tests.utils import get_mock_metadata, get_mock_study_metadata, ITEM_COUNT, MOCK_ENV
+from tests.utils import ITEM_COUNT, MOCK_ENV, get_mock_metadata, get_mock_study_metadata
 
 
-def _init_mock_data(s3_client, bucket_name, study, data_package):
+def _init_mock_data(s3_client, bucket_name, study, data_package, version):
     """Creates data in bucket for use in unit tests
 
     The following items are added:
@@ -51,13 +49,13 @@ def _init_mock_data(s3_client, bucket_name, study, data_package):
         "./tests/test_data/count_synthea_patient_agg.parquet",
         bucket_name,
         f"{BucketPath.AGGREGATE.value}/{study}/"
-        f"{study}__{data_package}/{study}__{data_package}__aggregate.parquet",
+        f"{study}__{data_package}/{version}/{study}__{data_package}__aggregate.parquet",
     )
     s3_client.upload_file(
         "./tests/test_data/count_synthea_patient_agg.csv",
         bucket_name,
         f"{BucketPath.CSVAGGREGATE.value}/{study}/"
-        f"{study}__{data_package}/{study}__{data_package}__aggregate.csv",
+        f"{study}__{data_package}/{version}/{study}__{data_package}__aggregate.csv",
     )
     s3_client.upload_file(
         "./tests/test_data/data_packages_cache.json",
@@ -90,8 +88,8 @@ def mock_bucket():
     bucket = os.environ["BUCKET_NAME"]
     s3_client.create_bucket(Bucket=bucket)
     aggregate_params = [
-        ["study", "encounter"],
-        ["other_study", "encounter"],
+        ["study", "encounter", "099"],
+        ["other_study", "encounter", "099"],
     ]
     for param_list in aggregate_params:
         _init_mock_data(s3_client, bucket, *param_list)
diff --git a/tests/dashboard/test_get_chart_data.py b/tests/dashboard/test_get_chart_data.py
index 643afe9..782729a 100644
--- a/tests/dashboard/test_get_chart_data.py
+++ b/tests/dashboard/test_get_chart_data.py
@@ -1,11 +1,10 @@
 import json
 import os
+from unittest import mock
 
 import pandas
 import pytest
 
-from unittest import mock
-
 from src.handlers.dashboard import get_chart_data
 from tests.utils import MOCK_ENV, TEST_BUCKET, TEST_GLUE_DB, TEST_WORKGROUP
 
diff --git a/tests/dashboard/test_get_metadata.py b/tests/dashboard/test_get_metadata.py
index 73084a9..b77b2fa 100644
--- a/tests/dashboard/test_get_metadata.py
+++ b/tests/dashboard/test_get_metadata.py
@@ -1,34 +1,44 @@
-import boto3
 import json
 import os
-
-import pytest
 from datetime import datetime, timezone
 from unittest import mock
 
+import boto3
+import pytest
+
+from src.handlers.dashboard.get_metadata import metadata_handler
 from src.handlers.shared.enums import BucketPath
 from src.handlers.shared.functions import read_metadata, write_metadata
-from src.handlers.dashboard.get_metadata import metadata_handler
-from tests.utils import get_mock_metadata, TEST_BUCKET
+from tests.utils import TEST_BUCKET, get_mock_metadata
 
 
 @pytest.mark.parametrize(
     "params,status,expected",
     [
         (None, 200, get_mock_metadata()),
-        ({"site": "general_hospital"}, 200, get_mock_metadata()["general_hospital"]),
         (
-            {"site": "general_hospital", "study": "study"},
+            {"site": "princeton_plainsboro_teaching_hospital"},
+            200,
+            get_mock_metadata()["princeton_plainsboro_teaching_hospital"],
+        ),
+        (
+            {"site": "princeton_plainsboro_teaching_hospital", "study": "study"},
             200,
-            get_mock_metadata()["general_hospital"]["study"],
+            get_mock_metadata()["princeton_plainsboro_teaching_hospital"]["study"],
         ),
         (
-            {"site": "general_hospital", "study": "study", "data_package": "encounter"},
+            {
+                "site": "princeton_plainsboro_teaching_hospital",
+                "study": "study",
+                "data_package": "encounter",
+            },
             200,
-            get_mock_metadata()["general_hospital"]["study"]["encounter"],
+            get_mock_metadata()["princeton_plainsboro_teaching_hospital"]["study"][
+                "encounter"
+            ],
         ),
         ({"site": "chicago_hope", "study": "study"}, 500, None),
-        ({"site": "general_hospital", "study": "flu"}, 500, None),
+        ({"site": "princeton_plainsboro_teaching_hospital", "study": "flu"}, 500, None),
     ],
 )
 def test_get_metadata(mock_bucket, params, status, expected):
diff --git a/tests/dashboard/test_get_study_periods.py b/tests/dashboard/test_get_study_periods.py
index 7f02efb..35eab78 100644
--- a/tests/dashboard/test_get_study_periods.py
+++ b/tests/dashboard/test_get_study_periods.py
@@ -1,14 +1,14 @@
-import boto3
 import json
 import os
-
-import pytest
 from datetime import datetime, timezone
 from unittest import mock
 
+import boto3
+import pytest
+
+from src.handlers.dashboard.get_study_periods import study_periods_handler
 from src.handlers.shared.enums import BucketPath
 from src.handlers.shared.functions import read_metadata, write_metadata
-from src.handlers.dashboard.get_study_periods import study_periods_handler
 from tests.utils import get_mock_study_metadata
 
 
@@ -17,17 +17,19 @@
     [
         (None, 200, get_mock_study_metadata()),
         (
-            {"site": "general_hospital"},
+            {"site": "princeton_plainsboro_teaching_hospital"},
             200,
-            get_mock_study_metadata()["general_hospital"],
+            get_mock_study_metadata()["princeton_plainsboro_teaching_hospital"],
         ),
         (
-            {"site": "general_hospital", "study": "study"},
+            {"site": "princeton_plainsboro_teaching_hospital", "study": "study"},
             200,
-            get_mock_study_metadata()["general_hospital"]["study"],
+            get_mock_study_metadata()["princeton_plainsboro_teaching_hospital"][
+                "study"
+            ],
         ),
         ({"site": "chicago_hope", "study": "study"}, 500, None),
-        ({"site": "general_hospital", "study": "flu"}, 500, None),
+        ({"site": "princeton_plainsboro_teaching_hospital", "study": "flu"}, 500, None),
     ],
 )
 def test_get_study_periods(mock_bucket, params, status, expected):
diff --git a/tests/dashboard/test_get_subscriptions.py b/tests/dashboard/test_get_subscriptions.py
index 86ddc41..5b928ca 100644
--- a/tests/dashboard/test_get_subscriptions.py
+++ b/tests/dashboard/test_get_subscriptions.py
@@ -1,14 +1,12 @@
 import os
-
 from unittest import mock
 
 import awswrangler
 import pandas
-
 from pytest_mock import MockerFixture
 
 from src.handlers.dashboard.get_data_packages import data_packages_handler
-from tests.utils import get_mock_metadata, MOCK_ENV, DATA_PACKAGE_COUNT
+from tests.utils import DATA_PACKAGE_COUNT, MOCK_ENV, get_mock_metadata
 
 
 @mock.patch.dict(os.environ, MOCK_ENV)
diff --git a/tests/site_upload/test_api_gateway_authorizer.py b/tests/site_upload/test_api_gateway_authorizer.py
index 44c8ed1..94bdc82 100644
--- a/tests/site_upload/test_api_gateway_authorizer.py
+++ b/tests/site_upload/test_api_gateway_authorizer.py
@@ -1,14 +1,13 @@
 import json
 import os
-import pytest
-
 from contextlib import nullcontext as does_not_raise
 from unittest import mock
 
+import pytest
 from pytest_mock import MockerFixture
 
 from src.handlers.site_upload.api_gateway_authorizer import lambda_handler
-from tests.utils import get_mock_auth, TEST_BUCKET
+from tests.utils import TEST_BUCKET, get_mock_auth
 
 
 @pytest.mark.parametrize(
diff --git a/tests/site_upload/test_cache_api.py b/tests/site_upload/test_cache_api.py
index f07f945..836677b 100644
--- a/tests/site_upload/test_cache_api.py
+++ b/tests/site_upload/test_cache_api.py
@@ -1,13 +1,12 @@
 import os
-import pytest
-
 from unittest import mock
 
 import awswrangler
 import pandas
+import pytest
 
 from src.handlers.site_upload.cache_api import cache_api_handler
-from tests.utils import get_mock_data_packages_cache, MOCK_ENV
+from tests.utils import MOCK_ENV, get_mock_data_packages_cache
 
 
 def mock_data_packages(*args, **kwargs):
diff --git a/tests/site_upload/test_fetch_upload_url.py b/tests/site_upload/test_fetch_upload_url.py
index 7bc6e7f..665e79b 100644
--- a/tests/site_upload/test_fetch_upload_url.py
+++ b/tests/site_upload/test_fetch_upload_url.py
@@ -1,12 +1,18 @@
-import boto3
 import json
 import os
-import pytest
-
 from unittest import mock
 
+import boto3
+import pytest
+
 from src.handlers.site_upload.fetch_upload_url import upload_url_handler
-from tests.utils import TEST_BUCKET, get_mock_metadata
+from tests.utils import (
+    EXISTING_DATA_P,
+    EXISTING_STUDY,
+    EXISTING_VERSION,
+    TEST_BUCKET,
+    get_mock_metadata,
+)
 
 
 @pytest.mark.parametrize(
@@ -14,9 +20,10 @@
     [
         (
             {
-                "study": "covid",
-                "data_package": "encounter",
+                "study": EXISTING_STUDY,
+                "data_package": EXISTING_DATA_P,
                 "filename": "encounter.parquet",
+                "version": EXISTING_VERSION,
             },
             200,
         ),
diff --git a/tests/site_upload/test_powerset_merge.py b/tests/site_upload/test_powerset_merge.py
index 8905087..1b6cca9 100644
--- a/tests/site_upload/test_powerset_merge.py
+++ b/tests/site_upload/test_powerset_merge.py
@@ -1,98 +1,112 @@
-import boto3
 import io
 import os
-
 from contextlib import nullcontext as does_not_raise
+from datetime import datetime, timezone
 from unittest import mock
 
 import awswrangler
+import boto3
 import pytest
-
-from datetime import datetime, timezone
-from pandas import DataFrame, read_parquet
 from freezegun import freeze_time
+from pandas import DataFrame, read_parquet
 
 from src.handlers.shared.enums import BucketPath
 from src.handlers.shared.functions import read_metadata, write_metadata
 from src.handlers.site_upload.powerset_merge import (
-    powerset_merge_handler,
-    expand_and_concat_sets,
     MergeError,
+    expand_and_concat_sets,
+    powerset_merge_handler,
+)
+from tests.utils import (
+    EXISTING_DATA_P,
+    EXISTING_SITE,
+    EXISTING_STUDY,
+    EXISTING_VERSION,
+    ITEM_COUNT,
+    MOCK_ENV,
+    NEW_DATA_P,
+    NEW_SITE,
+    NEW_STUDY,
+    NEW_VERSION,
+    OTHER_SITE,
+    OTHER_STUDY,
+    TEST_BUCKET,
+    get_mock_metadata,
 )
-
-from tests.utils import get_mock_metadata, TEST_BUCKET, ITEM_COUNT, MOCK_ENV
-
-
-SITE_NAME = "princeton_plainsboro_teaching_hospital"
-NEW_SITE_NAME = "chicago_hope"
-NEW_STUDY_NAME = "new_study"
-EXISTING_STUDY_NAME = "study"
-DATA_P_NAME = "encounter"
 
 
 @freeze_time("2020-01-01")
 @pytest.mark.parametrize(
-    "site,upload_file,upload_path,event_key,archives,status,expected_contents",
+    "upload_file,upload_path,event_key,archives,status,expected_contents",
     [
         (  # Adding a new data package to a site with uploads
-            f"{SITE_NAME}",
             "./tests/test_data/count_synthea_patient.parquet",
-            f"/{NEW_STUDY_NAME}/{DATA_P_NAME}/{SITE_NAME}/encounter.parquet",
-            f"/{NEW_STUDY_NAME}/{DATA_P_NAME}/{SITE_NAME}/encounter.parquet",
+            f"/{NEW_STUDY}/{EXISTING_DATA_P}/{EXISTING_SITE}/{EXISTING_VERSION}/encounter.parquet",
+            f"/{NEW_STUDY}/{EXISTING_DATA_P}/{EXISTING_SITE}/{EXISTING_VERSION}/encounter.parquet",
             False,
             200,
             ITEM_COUNT + 3,
         ),
         (  # Adding a new data package to a site without uploads
-            f"{NEW_SITE_NAME}",
             "./tests/test_data/count_synthea_patient.parquet",
-            f"/{NEW_STUDY_NAME}/{DATA_P_NAME}/{NEW_SITE_NAME}/encounter.parquet",
-            f"/{NEW_STUDY_NAME}/{DATA_P_NAME}/{NEW_SITE_NAME}/encounter.parquet",
+            f"/{NEW_STUDY}/{EXISTING_DATA_P}/{NEW_SITE}/{EXISTING_VERSION}/encounter.parquet",
+            f"/{NEW_STUDY}/{EXISTING_DATA_P}/{NEW_SITE}/{EXISTING_VERSION}/encounter.parquet",
             False,
             200,
             ITEM_COUNT + 3,
         ),
         (  # Updating an existing data package
-            f"{SITE_NAME}",
             "./tests/test_data/count_synthea_patient.parquet",
-            f"/{EXISTING_STUDY_NAME}/{DATA_P_NAME}/{SITE_NAME}/encounter.parquet",
-            f"/{EXISTING_STUDY_NAME}/{DATA_P_NAME}/{SITE_NAME}/encounter.parquet",
+            f"/{EXISTING_STUDY}/{EXISTING_DATA_P}/{EXISTING_SITE}/{EXISTING_VERSION}/encounter.parquet",
+            f"/{EXISTING_STUDY}/{EXISTING_DATA_P}/{EXISTING_SITE}/{EXISTING_VERSION}/encounter.parquet",
             True,
             200,
             ITEM_COUNT + 2,
         ),
+        (  # Updating an existing data package
+            "./tests/test_data/count_synthea_patient.parquet",
+            f"/{EXISTING_STUDY}/{EXISTING_DATA_P}/{EXISTING_SITE}/{NEW_VERSION}/encounter.parquet",
+            f"/{EXISTING_STUDY}/{EXISTING_DATA_P}/{EXISTING_SITE}/{NEW_VERSION}/encounter.parquet",
+            True,
+            200,
+            ITEM_COUNT + 4,
+        ),
         (  # Invalid parquet file
-            f"{SITE_NAME}",
             "./tests/site_upload/test_powerset_merge.py",
-            f"/{NEW_STUDY_NAME}/{DATA_P_NAME}/{SITE_NAME}/patient.parquet",
-            f"/{NEW_STUDY_NAME}/{DATA_P_NAME}/{SITE_NAME}/patient.parquet",
+            f"/{NEW_STUDY}/{EXISTING_DATA_P}/{EXISTING_SITE}/{EXISTING_VERSION}/patient.parquet",
+            f"/{NEW_STUDY}/{EXISTING_DATA_P}/{EXISTING_SITE}/{EXISTING_VERSION}/patient.parquet",
             False,
             500,
             ITEM_COUNT + 1,
         ),
         (  # Checks presence of commas in strings does not cause an error
-            f"{SITE_NAME}",
             "./tests/test_data/cube_strings_with_commas.parquet",
-            f"/{NEW_STUDY_NAME}/{DATA_P_NAME}/{SITE_NAME}/encounter.parquet",
-            f"/{NEW_STUDY_NAME}/{DATA_P_NAME}/{SITE_NAME}/encounter.parquet",
+            f"/{NEW_STUDY}/{EXISTING_DATA_P}/{EXISTING_SITE}/{EXISTING_VERSION}/encounter.parquet",
+            f"/{NEW_STUDY}/{EXISTING_DATA_P}/{EXISTING_SITE}/{EXISTING_VERSION}/encounter.parquet",
             False,
             200,
             ITEM_COUNT + 3,
         ),
         (  # Empty file upload
-            f"{SITE_NAME}",
             None,
-            f"/{NEW_STUDY_NAME}/{DATA_P_NAME}/{SITE_NAME}/encounter.parquet",
-            f"/{NEW_STUDY_NAME}/{DATA_P_NAME}/{SITE_NAME}/encounter.parquet",
+            f"/{NEW_STUDY}/{EXISTING_DATA_P}/{EXISTING_SITE}/{EXISTING_VERSION}/encounter.parquet",
+            f"/{NEW_STUDY}/{EXISTING_DATA_P}/{EXISTING_SITE}/{EXISTING_VERSION}/encounter.parquet",
             False,
             500,
             ITEM_COUNT + 1,
         ),
+        (  # Race condition - file deleted before job starts
+            None,
+            None,
+            f"/{NEW_STUDY}/{EXISTING_DATA_P}/{EXISTING_SITE}/{EXISTING_VERSION}/encounter.parquet",
+            False,
+            500,
+            ITEM_COUNT,
+        ),
     ],
 )
 @mock.patch.dict(os.environ, MOCK_ENV)
 def test_powerset_merge_single_upload(
-    site,
     upload_file,
     upload_path,
     event_key,
@@ -109,7 +123,7 @@ def test_powerset_merge_single_upload(
             TEST_BUCKET,
             f"{BucketPath.LATEST.value}{upload_path}",
         )
-    else:
+    elif upload_path is not None:
         with io.BytesIO(DataFrame().to_parquet()) as upload_fileobj:
             s3_client.upload_fileobj(
                 upload_fileobj,
@@ -133,9 +147,10 @@ def test_powerset_merge_single_upload(
     # This array looks like:
     # ['', 'study', 'package', 'site', 'file']
     event_list = event_key.split("/")
-    expected_study = event_list[1]
-    expected_package = event_list[2]
-    expected_site = event_list[3]
+    study = event_list[1]
+    package = event_list[2]
+    site = event_list[3]
+    version = event_list[4]
     res = powerset_merge_handler(event, {})
     assert res["statusCode"] == status
     s3_res = s3_client.list_objects_v2(Bucket=TEST_BUCKET)
@@ -144,34 +159,29 @@ def test_powerset_merge_single_upload(
         if item["Key"].endswith("aggregate.parquet"):
             assert item["Key"].startswith(BucketPath.AGGREGATE.value)
             # This finds the aggregate that was created/updated - ie it skips mocks
-            if (
-                expected_study in item["Key"]
-                and expected_study in item["Key"]
-                and status == 200
-            ):
+            if study in item["Key"] and status == 200:
                 agg_df = awswrangler.s3.read_parquet(
                     f"s3://{TEST_BUCKET}/{item['Key']}"
                 )
-                assert (agg_df["site"].eq(expected_site)).any()
+                assert (agg_df["site"].eq(site)).any()
         elif item["Key"].endswith("aggregate.csv"):
             assert item["Key"].startswith(BucketPath.CSVAGGREGATE.value)
         elif item["Key"].endswith("transactions.json"):
             assert item["Key"].startswith(BucketPath.META.value)
             metadata = read_metadata(s3_client, TEST_BUCKET)
             if res["statusCode"] == 200:
-                study = event_key.split("/")[1]
                 assert (
-                    metadata[site][study][DATA_P_NAME]["last_aggregation"]
+                    metadata[site][study][EXISTING_DATA_P][version]["last_aggregation"]
                     == datetime.now(timezone.utc).isoformat()
                 )
             else:
                 assert (
-                    metadata["general_hospital"]["study"]["encounter"][
-                        "last_aggregation"
-                    ]
-                    == get_mock_metadata()["general_hospital"]["study"]["encounter"][
-                        "last_aggregation"
-                    ]
+                    metadata["princeton_plainsboro_teaching_hospital"]["study"][
+                        "encounter"
+                    ]["099"]["last_aggregation"]
+                    == get_mock_metadata()["princeton_plainsboro_teaching_hospital"][
+                        "study"
+                    ]["encounter"]["099"]["last_aggregation"]
                 )
         elif item["Key"].startswith(BucketPath.LAST_VALID.value):
             assert item["Key"] == (f"{BucketPath.LAST_VALID.value}{upload_path}")
@@ -214,31 +224,31 @@ def test_powerset_merge_join_study_data(
     s3_client.upload_file(
         upload_file,
         TEST_BUCKET,
-        f"{BucketPath.LATEST.value}/{EXISTING_STUDY_NAME}/"
-        f"{DATA_P_NAME}/{NEW_SITE_NAME}/encounter.parquet",
+        f"{BucketPath.LATEST.value}/{EXISTING_STUDY}/"
+        f"{EXISTING_DATA_P}/{NEW_SITE}/{EXISTING_VERSION}/encounter.parquet",
     )
 
     s3_client.upload_file(
         "./tests/test_data/count_synthea_patient.parquet",
         TEST_BUCKET,
-        f"{BucketPath.LAST_VALID.value}/{EXISTING_STUDY_NAME}/"
-        f"{DATA_P_NAME}/{SITE_NAME}/encounter.parquet",
+        f"{BucketPath.LAST_VALID.value}/{EXISTING_STUDY}/"
+        f"{EXISTING_DATA_P}/{EXISTING_SITE}/{EXISTING_VERSION}/encounter.parquet",
     )
 
     if archives:
         s3_client.upload_file(
             "./tests/test_data/count_synthea_patient.parquet",
             TEST_BUCKET,
-            f"{BucketPath.LAST_VALID.value}/{EXISTING_STUDY_NAME}/"
-            f"{DATA_P_NAME}/{NEW_SITE_NAME}/encounter.parquet",
+            f"{BucketPath.LAST_VALID.value}/{EXISTING_STUDY}/"
+            f"{EXISTING_DATA_P}/{NEW_SITE}/{EXISTING_VERSION}/encounter.parquet",
         )
 
     event = {
         "Records": [
             {
                 "Sns": {
-                    "Message": f"{BucketPath.LATEST.value}/{EXISTING_STUDY_NAME}"
-                    f"/{DATA_P_NAME}/{NEW_SITE_NAME}/encounter.parquet"
+                    "Message": f"{BucketPath.LATEST.value}/{EXISTING_STUDY}"
+                    f"/{EXISTING_DATA_P}/{NEW_SITE}/{EXISTING_VERSION}/encounter.parquet"
                 },
             }
         ]
@@ -289,4 +299,4 @@ def test_expand_and_concat(mock_bucket, upload_file, load_empty, raises):
             TEST_BUCKET,
             s3_path,
         )
-        expand_and_concat_sets(df, f"s3://{TEST_BUCKET}/{s3_path}", EXISTING_STUDY_NAME)
+        expand_and_concat_sets(df, f"s3://{TEST_BUCKET}/{s3_path}", EXISTING_STUDY)
diff --git a/tests/site_upload/test_process_upload.py b/tests/site_upload/test_process_upload.py
index 531c185..34ad19f 100644
--- a/tests/site_upload/test_process_upload.py
+++ b/tests/site_upload/test_process_upload.py
@@ -1,72 +1,84 @@
-import boto3
 import os
+from datetime import datetime, timezone
 
+import boto3
 import pytest
-
-from datetime import datetime, timezone
 from freezegun import freeze_time
 
 from src.handlers.shared.enums import BucketPath
 from src.handlers.shared.functions import read_metadata, write_metadata
 from src.handlers.site_upload.process_upload import process_upload_handler
-
-from tests.utils import TEST_BUCKET, ITEM_COUNT
+from tests.utils import (
+    EXISTING_DATA_P,
+    EXISTING_SITE,
+    EXISTING_STUDY,
+    EXISTING_VERSION,
+    ITEM_COUNT,
+    NEW_DATA_P,
+    NEW_SITE,
+    NEW_STUDY,
+    NEW_VERSION,
+    OTHER_SITE,
+    OTHER_STUDY,
+    TEST_BUCKET,
+)
 
 
 @freeze_time("2020-01-01")
 @pytest.mark.parametrize(
-    "site,upload_file,upload_path,event_key,status,expected_contents",
+    "upload_file,upload_path,event_key,status,expected_contents",
     [
         (  # Adding a new data package to a site with uploads
-            "princeton_plainsboro_teaching_hospital",
             "./tests/test_data/cube_simple_example.parquet",
-            "/covid/encounter/princeton_plainsboro_teaching_hospital/document.parquet",
-            "/covid/encounter/princeton_plainsboro_teaching_hospital/document.parquet",
+            f"/{EXISTING_STUDY}/{NEW_DATA_P}/{EXISTING_SITE}/{EXISTING_VERSION}/document.parquet",
+            f"/{EXISTING_STUDY}/{NEW_DATA_P}/{EXISTING_SITE}/{EXISTING_VERSION}/document.parquet",
             200,
             ITEM_COUNT + 1,
         ),
         (  # Adding a new data package to a site without uploads
-            "chicago_hope",
             "./tests/test_data/cube_simple_example.parquet",
-            "/covid/encounter/chicago_hope/document.parquet",
-            "/covid/encounter/chicago_hope/document.parquet",
+            f"/{EXISTING_STUDY}/{NEW_DATA_P}/{NEW_SITE}/{EXISTING_VERSION}/document.parquet",
+            f"/{EXISTING_STUDY}/{NEW_DATA_P}/{NEW_SITE}/{EXISTING_VERSION}/document.parquet",
             200,
             ITEM_COUNT + 1,
         ),
         (  # Updating an existing data package
-            "princeton_plainsboro_teaching_hospital",
             "./tests/test_data/cube_simple_example.parquet",
-            "/covid/encounter/princeton_plainsboro_teaching_hospital/encounter.parquet",
-            "/covid/encounter/princeton_plainsboro_teaching_hospital/encounter.parquet",
+            f"/{EXISTING_STUDY}/{EXISTING_DATA_P}/{EXISTING_SITE}/{EXISTING_VERSION}/encounter.parquet",
+            f"/{EXISTING_STUDY}/{EXISTING_DATA_P}/{EXISTING_SITE}/{EXISTING_VERSION}/encounter.parquet",
+            200,
+            ITEM_COUNT + 1,
+        ),
+        (  # New version of an existing data package
+            "./tests/test_data/cube_simple_example.parquet",
+            f"/{EXISTING_STUDY}/{EXISTING_DATA_P}/{EXISTING_SITE}/{NEW_VERSION}/encounter.parquet",
+            f"/{EXISTING_STUDY}/{EXISTING_DATA_P}/{EXISTING_SITE}/{NEW_VERSION}/encounter.parquet",
             200,
             ITEM_COUNT + 1,
         ),
         (  # Non-parquet file
-            "princeton_plainsboro_teaching_hospital",
             "./tests/test_data/cube_simple_example.csv",
-            "/covid/encounter/princeton_plainsboro_teaching_hospital/document.csv",
-            "/covid/encounter/princeton_plainsboro_teaching_hospital/document.csv",
+            f"/{EXISTING_STUDY}/{NEW_DATA_P}/{EXISTING_SITE}/{EXISTING_VERSION}/document.csv",
+            f"/{EXISTING_STUDY}/{NEW_DATA_P}/{EXISTING_SITE}/{EXISTING_VERSION}/document.csv",
             500,
             ITEM_COUNT + 1,
         ),
         (  # S3 event dispatched when file is not present
-            "princeton_plainsboro_teaching_hospital",
             None,
             None,
-            "/covid/encounter/princeton_plainsboro_teaching_hospital/missing.parquet",
+            f"/{EXISTING_STUDY}/{EXISTING_DATA_P}/{EXISTING_SITE}/{EXISTING_VERSION}/missing.parquet",
             500,
             ITEM_COUNT,
         ),
-        (  # Adding study metadata data package
-            "princeton_plainsboro_teaching_hospital",
+        (  # Adding metadata data package
             "./tests/test_data/cube_simple_example.parquet",
             (
-                "/covid/encounter/princeton_plainsboro_teaching_hospital/"
-                "document_meta_date.parquet"
+                f"/{EXISTING_STUDY}/{EXISTING_DATA_P}/{EXISTING_SITE}/"
+                f"{EXISTING_VERSION}/document_meta_date.parquet"
             ),
             (
-                "/covid/encounter/princeton_plainsboro_teaching_hospital/"
-                "document_meta_date.parquet"
+                f"/{EXISTING_STUDY}/{EXISTING_DATA_P}/{EXISTING_SITE}/"
+                f"{EXISTING_VERSION}/document_meta_date.parquet"
             ),
             200,
             ITEM_COUNT + 1,
@@ -74,7 +86,6 @@
     ],
 )
 def test_process_upload(
-    site,
     upload_file,
     upload_path,
     event_key,
@@ -111,9 +122,14 @@ def test_process_upload(
         elif item["Key"].endswith("transactions.json"):
             assert item["Key"].startswith(BucketPath.META.value)
             metadata = read_metadata(s3_client, TEST_BUCKET)
-            if upload_file is not None:
+            if upload_file is not None and upload_path is not None:
+                path_params = upload_path.split("/")
+                study = path_params[1]
+                data_package = path_params[2]
+                site = path_params[3]
+                version = path_params[4]
                 assert (
-                    metadata[site]["covid"]["encounter"]["last_upload"]
+                    metadata[site][study][data_package][version]["last_upload"]
                     == datetime.now(timezone.utc).isoformat()
                 )
         elif item["Key"].startswith(BucketPath.STUDY_META.value):
diff --git a/tests/site_upload/test_study_period.py b/tests/site_upload/test_study_period.py
index 0f78481..492b36a 100644
--- a/tests/site_upload/test_study_period.py
+++ b/tests/site_upload/test_study_period.py
@@ -1,78 +1,92 @@
-import boto3
 import csv
 import os
+from datetime import datetime, timezone
 
+import boto3
 import pytest
-from datetime import datetime, timezone
 from freezegun import freeze_time
 
 from src.handlers.shared.enums import BucketPath
 from src.handlers.shared.functions import read_metadata, write_metadata
 from src.handlers.site_upload.study_period import study_period_handler
-
-from tests.utils import get_mock_study_metadata, TEST_BUCKET
+from tests.utils import (
+    EXISTING_DATA_P,
+    EXISTING_SITE,
+    EXISTING_STUDY,
+    EXISTING_VERSION,
+    NEW_DATA_P,
+    NEW_SITE,
+    NEW_STUDY,
+    NEW_VERSION,
+    OTHER_SITE,
+    OTHER_STUDY,
+    TEST_BUCKET,
+    get_mock_study_metadata,
+)
 
 
 @freeze_time("2020-01-01")
 @pytest.mark.parametrize(
-    "site,upload_file,upload_path,event_key,status,study_key",
+    "upload_file,upload_path,event_key,status",
     [
         (  # Adding a new study to an existing site
-            "princeton_plainsboro_teaching_hospital",
             "./tests/test_data/meta_date.parquet",
             (
-                "/test/test_meta_date/princeton_plainsboro_teaching_hospital/"
+                f"/{NEW_STUDY}/{EXISTING_DATA_P}/{EXISTING_SITE}/{EXISTING_VERSION}/"
                 "test_meta_date.parquet"
             ),
             (
-                "/test/test_meta_date/princeton_plainsboro_teaching_hospital/"
+                f"/{NEW_STUDY}/{EXISTING_DATA_P}/{EXISTING_SITE}/{EXISTING_VERSION}/"
                 "test_meta_date.parquet"
             ),
             200,
-            "test",
         ),
         (  # Adding a new study to a new site
-            "chicago_hope",
             "./tests/test_data/meta_date.parquet",
-            "/test/test_meta_date/chicago_hope/test_meta_date.parquet",
-            "/test/test_meta_date/chicago_hope/test_meta_date.parquet",
+            f"/{NEW_STUDY}/{EXISTING_DATA_P}/{NEW_SITE}/{EXISTING_VERSION}/test_meta_date.parquet",
+            f"/{NEW_STUDY}/{EXISTING_DATA_P}/{NEW_SITE}/{EXISTING_VERSION}/test_meta_date.parquet",
+            200,
+        ),
+        (  # newer version of existing study
+            "./tests/test_data/meta_date.parquet",
+            (
+                f"/{EXISTING_STUDY}/{EXISTING_DATA_P}/{EXISTING_SITE}/{NEW_VERSION}/"
+                "test_meta_date.parquet"
+            ),
+            (
+                f"/{EXISTING_STUDY}/{EXISTING_DATA_P}/{EXISTING_SITE}/{NEW_VERSION}/"
+                "test_meta_date.parquet"
+            ),
             200,
-            "test",
         ),
         (  # updating an existing study
-            "princeton_plainsboro_teaching_hospital",
             "./tests/test_data/meta_date.parquet",
             (
-                "/covid/test_meta_date/princeton_plainsboro_teaching_hospital/"
+                f"/{EXISTING_STUDY}/{EXISTING_DATA_P}/{EXISTING_SITE}/{EXISTING_VERSION}/"
                 "test_meta_date.parquet"
             ),
             (
-                "/covid/test_meta_date/princeton_plainsboro_teaching_hospital/"
+                f"/{EXISTING_STUDY}/{EXISTING_DATA_P}/{EXISTING_SITE}/{EXISTING_VERSION}/"
                 "test_meta_date.parquet"
             ),
             200,
-            "covid",
         ),
         (  # invalid file
-            "princeton_plainsboro_teaching_hospital",
             "./tests/test_data/meta_date.parquet",
             None,
             (
-                "/covid/test_meta_date/princeton_plainsboro_teaching_hospital/"
+                f"/{EXISTING_STUDY}/{EXISTING_DATA_P}/{EXISTING_SITE}/{EXISTING_VERSION}/"
                 "wrong.parquet"
             ),
             500,
-            None,
         ),
     ],
 )
 def test_process_upload(
-    site,
     upload_file,
     upload_path,
     event_key,
     status,
-    study_key,
     mock_bucket,
 ):
     s3_client = boto3.client("s3", region_name="us-east-1")
@@ -88,10 +102,14 @@ def test_process_upload(
     res = study_period_handler(event, {})
     assert res["statusCode"] == status
     metadata = read_metadata(s3_client, TEST_BUCKET, meta_type="study_periods")
-    if study_key is not None:
-        assert study_key in metadata[site]
+    if upload_file is not None and upload_path is not None:
+        path_params = upload_path.split("/")
+        study = path_params[1]
+        site = path_params[3]
+        version = path_params[4]
+        assert study in metadata[site]
         assert (
-            metadata[site][study_key]["last_data_update"]
+            metadata[site][study][version]["last_data_update"]
             == datetime.now(timezone.utc).isoformat()
         )
         with open("./tests/test_data/meta_date.csv", "r") as file:
@@ -99,5 +117,7 @@ def test_process_upload(
             # discarding CSV header row
             next(reader)
             row = next(reader)
-            assert metadata[site][study_key]["earliest_date"] == f"{row[0]}T00:00:00"
-            assert metadata[site][study_key]["latest_date"] == f"{row[1]}T00:00:00"
+            assert (
+                metadata[site][study][version]["earliest_date"] == f"{row[0]}T00:00:00"
+            )
+            assert metadata[site][study][version]["latest_date"] == f"{row[1]}T00:00:00"
diff --git a/tests/utils.py b/tests/utils.py
index 6eb0a42..8f479cf 100644
--- a/tests/utils.py
+++ b/tests/utils.py
@@ -9,6 +9,17 @@
 ITEM_COUNT = 9
 DATA_PACKAGE_COUNT = 2
 
+EXISTING_SITE = "princeton_plainsboro_teaching_hospital"
+NEW_SITE = "chicago_hope"
+OTHER_SITE = "st_elsewhere"
+NEW_STUDY = "new_study"
+EXISTING_STUDY = "study"
+OTHER_STUDY = "other_study"
+EXISTING_DATA_P = "encounter"
+NEW_DATA_P = "document"
+EXISTING_VERSION = "099"
+NEW_VERSION = "100"
+
 # This is a convenience for loading into os.environ with mock.patch.dict.
 # Other cases should probably use the getter version below.
 MOCK_ENV = {
@@ -23,37 +34,43 @@
 
 def get_mock_metadata():
     return {
-        "general_hospital": {
-            "study": {
-                "encounter": {
-                    "version": "1.0",
-                    "last_upload": "2023-02-24T15:03:34+00:00",
-                    "last_data_update": "2023-02-24T15:03:40.657583+00:00",
-                    "last_aggregation": "2023-02-24T15:08:07.504595+00:00",
-                    "last_error": None,
-                    "deleted": None,
+        EXISTING_SITE: {
+            EXISTING_STUDY: {
+                EXISTING_DATA_P: {
+                    EXISTING_VERSION: {
+                        "version": "1.0",
+                        "last_upload": "2023-02-24T15:03:34+00:00",
+                        "last_data_update": "2023-02-24T15:03:40.657583+00:00",
+                        "last_aggregation": "2023-02-24T15:08:07.504595+00:00",
+                        "last_error": None,
+                        "deleted": None,
+                    }
                 }
             },
-            "other_study": {
-                "encounter": {
-                    "version": "1.0",
-                    "last_upload": "2023-02-24T15:43:57+00:00",
-                    "last_data_update": "2023-02-24T15:44:03.861574+00:00",
-                    "last_aggregation": "2023-02-24T15:44:03.861574+00:00",
-                    "last_error": None,
-                    "deleted": None,
+            OTHER_STUDY: {
+                EXISTING_DATA_P: {
+                    EXISTING_VERSION: {
+                        "version": "1.0",
+                        "last_upload": "2023-02-24T15:43:57+00:00",
+                        "last_data_update": "2023-02-24T15:44:03.861574+00:00",
+                        "last_aggregation": "2023-02-24T15:44:03.861574+00:00",
+                        "last_error": None,
+                        "deleted": None,
+                    }
                 }
             },
         },
-        "st_elsewhere": {
-            "study": {
-                "encounter": {
-                    "version": "1.0",
-                    "last_upload": "2023-02-24T15:08:06+00:00",
-                    "last_data_update": "2023-02-24T15:08:07.771080+00:00",
-                    "last_aggregation": "2023-02-24T15:08:07.771080+00:00",
-                    "last_error": None,
-                    "deleted": None,
+        OTHER_SITE: {
+            EXISTING_STUDY: {
+                EXISTING_DATA_P: {
+                    EXISTING_VERSION: {
+                        "version": "1.0",
+                        "last_upload": "2023-02-24T15:08:06+00:00",
+                        "last_data_update": "2023-02-24T15:08:07.771080+00:00",
+                        "last_aggregation": "2023-02-24T15:08:07.771080+00:00",
+                        "last_error": None,
+                        "deleted": None,
+                    }
                 }
             }
         },
@@ -62,26 +79,32 @@ def get_mock_metadata():
 
 def get_mock_study_metadata():
     return {
-        "general_hospital": {
-            "study": {
-                "version": "1.0",
-                "last_data_update": "2023-02-24T15:03:40.657583+00:00",
-                "earliest_data": "2020-02-24T15:03:40.657583+00:00",
-                "latest_data": "2023-02-24T15:03:40.657583+00:00",
+        EXISTING_SITE: {
+            EXISTING_STUDY: {
+                EXISTING_VERSION: {
+                    "version": "1.0",
+                    "last_data_update": "2023-02-24T15:03:40.657583+00:00",
+                    "earliest_data": "2020-02-24T15:03:40.657583+00:00",
+                    "latest_data": "2023-02-24T15:03:40.657583+00:00",
+                }
             },
-            "other_study": {
-                "version": "1.0",
-                "last_data_update": "2023-02-24T15:44:03.861574+00:00",
-                "earliest_data": "2020-02-24T15:03:40.657583+00:00",
-                "latest_data": "2023-02-24T15:03:40.657583+00:00",
+            OTHER_STUDY: {
+                EXISTING_VERSION: {
+                    "version": "1.0",
+                    "last_data_update": "2023-02-24T15:44:03.861574+00:00",
+                    "earliest_data": "2020-02-24T15:03:40.657583+00:00",
+                    "latest_data": "2023-02-24T15:03:40.657583+00:00",
+                }
             },
         },
-        "st_elsewhere": {
-            "study": {
-                "version": "1.0",
-                "last_data_update": "2023-02-24T15:08:07.771080+00:00",
-                "earliest_data": "2020-02-24T15:03:40.657583+00:00",
-                "latest_data": "2023-02-24T15:03:40.657583+00:00",
+        OTHER_SITE: {
+            EXISTING_STUDY: {
+                EXISTING_VERSION: {
+                    "version": "1.0",
+                    "last_data_update": "2023-02-24T15:08:07.771080+00:00",
+                    "earliest_data": "2020-02-24T15:03:40.657583+00:00",
+                    "latest_data": "2023-02-24T15:03:40.657583+00:00",
+                }
             }
         },
     }