diff --git a/examples/21_tcp_echo_server.rs b/examples/21_tcp_echo_server.rs
new file mode 100644
index 0000000..0401972
--- /dev/null
+++ b/examples/21_tcp_echo_server.rs
@@ -0,0 +1,134 @@
+//! This example demonstrates how to gracefully shutdown a server
+//! that spawns an indefinite number of connection tasks.
+//!
+//! The server is a simple TCP echo server, capitalizing the data
+//! it echos (to demonstrate that it computes things).
+//! On shutdown, it transmits a goodbye message, to demonstrate
+//! that during shutdown we can still perform cleanup steps.
+//!
+//! This example is similar to the hyper example; for a more complex
+//! version of this same example, look there.
+
+use miette::{Context, IntoDiagnostic, Result};
+use tokio::io::{AsyncReadExt, AsyncWriteExt};
+use tokio::time::Duration;
+use tokio_graceful_shutdown::errors::CancelledByShutdown;
+use tokio_graceful_shutdown::{FutureExt, SubsystemBuilder, SubsystemHandle, Toplevel};
+
+use std::net::SocketAddr;
+
+use tokio::net::{TcpListener, TcpStream};
+use tokio_util::task::TaskTracker;
+
+async fn echo_connection(tcp: &mut TcpStream) -> Result<()> {
+    tcp.write_all(b"Hello!\r\n").await.into_diagnostic()?;
+
+    let mut buffer = [0u8; 256];
+    loop {
+        match tcp.read(&mut buffer).await {
+            Ok(0) => return Ok(()),
+            Err(e) => return Err(e).into_diagnostic(),
+            Ok(len) => {
+                let bytes = &mut buffer[..len];
+                for byte in bytes.iter_mut() {
+                    *byte = byte.to_ascii_uppercase();
+                }
+                tcp.write_all(bytes).await.into_diagnostic()?;
+            }
+        }
+    }
+}
+
+async fn echo_connection_shutdown(tcp: &mut TcpStream) -> Result<()> {
+    tcp.write_all(b"Goodbye.\r\n").await.into_diagnostic()?;
+    tcp.shutdown().await.into_diagnostic()?;
+
+    Ok(())
+}
+
+async fn connection_handler(
+    subsys: SubsystemHandle,
+    listener: TcpListener,
+    connection_tracker: TaskTracker,
+) -> Result<()> {
+    loop {
+        let connection = match listener.accept().cancel_on_shutdown(&subsys).await {
+            Ok(connection) => connection,
+            Err(CancelledByShutdown) => break,
+        };
+        let (mut tcp, addr) = connection
+            .into_diagnostic()
+            .context("Error while waiting for connection")?;
+
+        // Spawn handler on connection tracker to give the parent subsystem
+        // the chance to wait for the shutdown to finish
+        connection_tracker.spawn({
+            let cancellation_token = subsys.create_cancellation_token();
+            async move {
+                tracing::info!("Connected to {} ...", addr);
+
+                let result = tokio::select! {
+                    e = echo_connection(&mut tcp) => e,
+                    _ = cancellation_token.cancelled() => {
+                        tracing::info!("Shutting down {} ...", addr);
+                        echo_connection_shutdown(&mut tcp).await
+                    },
+                };
+
+                if let Err(err) = result {
+                    tracing::warn!("Error serving connection: {:?}", err);
+                } else {
+                    tracing::info!("Connection to {} closed.", addr);
+                }
+            }
+        });
+    }
+
+    Ok(())
+}
+
+async fn echo_subsystem(subsys: SubsystemHandle) -> Result<()> {
+    let addr: SocketAddr = ([127, 0, 0, 1], 12345).into();
+
+    // Bind to the port and listen for incoming TCP connections
+    let listener = TcpListener::bind(addr)
+        .await
+        .into_diagnostic()
+        .context("Unable to start tcp server")?;
+    tracing::info!("Listening on {}", addr);
+
+    // Use a tasktracker instead of spawning a subsystem for every connection,
+    // as this would result in a lot of overhead.
+    let connection_tracker = TaskTracker::new();
+
+    let listener = subsys.start(SubsystemBuilder::new("Echo Listener", {
+        let connection_tracker = connection_tracker.clone();
+        move |subsys| connection_handler(subsys, listener, connection_tracker)
+    }));
+
+    // Make sure no more tasks can be spawned before we close the tracker
+    listener.join().await?;
+
+    // Wait for connections to close
+    connection_tracker.close();
+    connection_tracker.wait().await;
+
+    Ok(())
+}
+
+#[tokio::main]
+async fn main() -> Result<()> {
+    // Init logging
+    tracing_subscriber::fmt()
+        .with_max_level(tracing::Level::DEBUG)
+        .init();
+
+    // Setup and execute subsystem tree
+    Toplevel::new(|s| async move {
+        s.start(SubsystemBuilder::new("EchoServer", echo_subsystem));
+    })
+    .catch_signals()
+    .handle_shutdown_requests(Duration::from_secs(5))
+    .await
+    .map_err(Into::into)
+}