diff --git a/CHANGELOG.md b/CHANGELOG.md
index 95d22a4600..b26122d09f 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -12,6 +12,8 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 ## Unreleased
 
 ### Fixed
+- `opentelemetry-instrumentation-flask` Fixed an issue where http_server_request_duration metrics was being recorded for excluded urls.
+  ([#3794](https://github.com/open-telemetry/opentelemetry-python-contrib/pull/3794))
 
 - `opentelemetry-instrumentation-dbapi`: fix crash retrieving libpq version when enabling commenter with psycopg
   ([#3796](https://github.com/open-telemetry/opentelemetry-python-contrib/pull/3796))
diff --git a/instrumentation/opentelemetry-instrumentation-flask/src/opentelemetry/instrumentation/flask/__init__.py b/instrumentation/opentelemetry-instrumentation-flask/src/opentelemetry/instrumentation/flask/__init__.py
index 9a22383e1b..2dfe135ff2 100644
--- a/instrumentation/opentelemetry-instrumentation-flask/src/opentelemetry/instrumentation/flask/__init__.py
+++ b/instrumentation/opentelemetry-instrumentation-flask/src/opentelemetry/instrumentation/flask/__init__.py
@@ -395,30 +395,34 @@ def _start_response(status, response_headers, *args, **kwargs):
             return start_response(status, response_headers, *args, **kwargs)
 
         result = wsgi_app(wrapped_app_environ, _start_response)
-        duration_s = default_timer() - start
-        if duration_histogram_old:
-            duration_attrs_old = otel_wsgi._parse_duration_attrs(
-                attributes, _StabilityMode.DEFAULT
-            )
+        if flask.request and (
+            excluded_urls is None
+            or not excluded_urls.url_disabled(flask.request.url)
+        ):
+            duration_s = default_timer() - start
+            if duration_histogram_old:
+                duration_attrs_old = otel_wsgi._parse_duration_attrs(
+                    attributes, _StabilityMode.DEFAULT
+                )
 
-            if request_route:
-                # http.target to be included in old semantic conventions
-                duration_attrs_old[HTTP_TARGET] = str(request_route)
+                if request_route:
+                    # http.target to be included in old semantic conventions
+                    duration_attrs_old[HTTP_TARGET] = str(request_route)
 
-            duration_histogram_old.record(
-                max(round(duration_s * 1000), 0), duration_attrs_old
-            )
-        if duration_histogram_new:
-            duration_attrs_new = otel_wsgi._parse_duration_attrs(
-                attributes, _StabilityMode.HTTP
-            )
+                duration_histogram_old.record(
+                    max(round(duration_s * 1000), 0), duration_attrs_old
+                )
+            if duration_histogram_new:
+                duration_attrs_new = otel_wsgi._parse_duration_attrs(
+                    attributes, _StabilityMode.HTTP
+                )
 
-            if request_route:
-                duration_attrs_new[HTTP_ROUTE] = str(request_route)
+                if request_route:
+                    duration_attrs_new[HTTP_ROUTE] = str(request_route)
 
-            duration_histogram_new.record(
-                max(duration_s, 0), duration_attrs_new
-            )
+                duration_histogram_new.record(
+                    max(duration_s, 0), duration_attrs_new
+                )
         active_requests_counter.add(-1, active_requests_count_attrs)
         return result
 
diff --git a/instrumentation/opentelemetry-instrumentation-flask/tests/test_programmatic.py b/instrumentation/opentelemetry-instrumentation-flask/tests/test_programmatic.py
index 60f08c7cd2..3ff2345297 100644
--- a/instrumentation/opentelemetry-instrumentation-flask/tests/test_programmatic.py
+++ b/instrumentation/opentelemetry-instrumentation-flask/tests/test_programmatic.py
@@ -523,7 +523,8 @@ def test_flask_metrics(self):
                                 attr,
                                 _recommended_metrics_attrs_old[metric.name],
                             )
-        self.assertTrue(number_data_point_seen and histogram_data_point_seen)
+        self.assertTrue(number_data_point_seen)
+        self.assertFalse(histogram_data_point_seen)
 
     def test_flask_metrics_new_semconv(self):
         start = default_timer()
@@ -561,7 +562,8 @@ def test_flask_metrics_new_semconv(self):
                                 attr,
                                 _recommended_metrics_attrs_new[metric.name],
                             )
-        self.assertTrue(number_data_point_seen and histogram_data_point_seen)
+        self.assertTrue(number_data_point_seen)
+        self.assertFalse(histogram_data_point_seen)
 
     def test_flask_metric_values(self):
         start = default_timer()