diff --git a/evolutionary_search/__init__.py b/evolutionary_search/__init__.py
index 60a2e6b..99e2652 100644
--- a/evolutionary_search/__init__.py
+++ b/evolutionary_search/__init__.py
@@ -1,2 +1,3 @@
 from .cv import *
 from .optimize import maximize, compile
+
diff --git a/evolutionary_search/cv.py b/evolutionary_search/cv.py
index 74d6a9f..b6bc8f2 100644
--- a/evolutionary_search/cv.py
+++ b/evolutionary_search/cv.py
@@ -216,6 +216,10 @@ class EvolutionaryAlgorithmSearchCV(BaseSearchCV):
         FitFailedWarning is raised. This parameter does not affect the refit
         step, which will always raise the error.
 
+    maximize : boolean, default=True
+        If True, maximize the scoring function result. Otherwise, minimize
+        the scoring function result.
+
 
     Examples
     --------
@@ -283,7 +287,7 @@ def __init__(self, estimator, params, scoring=None, cv=4,
                  gene_mutation_prob=0.1, gene_crossover_prob=0.5,
                  tournament_size=3, generations_number=10, gene_type=None,
                  n_jobs=1, iid=True, error_score='raise',
-                 fit_params={}):
+                 fit_params={}, maximize=True):
         super(EvolutionaryAlgorithmSearchCV, self).__init__(
             estimator=estimator, scoring=scoring, fit_params=fit_params,
             iid=iid, refit=refit, cv=cv, verbose=verbose,
@@ -302,8 +306,16 @@ def __init__(self, estimator, params, scoring=None, cv=4,
         self.best_params_ = None
         self.score_cache = {}
         self.n_jobs = n_jobs
-        creator.create("FitnessMax", base.Fitness, weights=(1.0,))
-        creator.create("Individual", list, est=clone(self.estimator), fitness=creator.FitnessMax)
+        self.maximize = maximize
+        if maximize:
+            weights = (1.0,)
+            fitness_name = "FitnessMax"
+        else:
+            weights = (-1.0,)
+            fitness_name = "FitnessMin"
+        creator.create(fitness_name, base.Fitness, weights=weights)
+        creator.create("Individual", list, est=clone(self.estimator),
+                       fitness=getattr(creator, fitness_name))
 
     @property
     def possible_params(self):
@@ -349,7 +361,10 @@ def best_index_(self):
 
     def fit(self, X, y=None):
         self.best_estimator_ = None
-        self.best_mem_score_ = float("-inf")
+        if self.maximize:
+            self.best_mem_score_ = float("-inf")
+        else:
+            self.best_mem_score_ = float("inf")
         self.best_mem_params_ = None
         for possible_params in self.possible_params:
             _check_param_grid(possible_params)
@@ -454,7 +469,8 @@ def _fit(self, X, y, parameter_dict):
             print("Best individual is: %s\nwith fitness: %s" % (
                 current_best_params_, current_best_score_))
 
-        if current_best_score_ > self.best_mem_score_:
+        if ((self.maximize and (current_best_score_ > self.best_mem_score_)) or
+            (not self.maximize and (current_best_score_ < self.best_mem_score_))):
             self.best_mem_score_ = current_best_score_
             self.best_mem_params_ = current_best_params_
 
@@ -468,3 +484,4 @@ def _fit(self, X, y, parameter_dict):
 
         self.best_score_ = current_best_score_
         self.best_params_ = current_best_params_
+