diff --git a/releasenotes/notes/add-get-recent-failed-runs-by-run-meta-f36c5197f3fe869f.yaml b/releasenotes/notes/add-get-recent-failed-runs-by-run-meta-f36c5197f3fe869f.yaml
new file mode 100644
index 0000000..4b76418
--- /dev/null
+++ b/releasenotes/notes/add-get-recent-failed-runs-by-run-meta-f36c5197f3fe869f.yaml
@@ -0,0 +1,5 @@
+---
+features:
+  - Adds a new DB API method get_recent_failed_runs_by_run_metadata() that
+    will return a list of models.Run objects for recent failures on runs
+    with a specified run_metadata key value pair
diff --git a/subunit2sql/db/api.py b/subunit2sql/db/api.py
index ad3fbe8..c064bf9 100644
--- a/subunit2sql/db/api.py
+++ b/subunit2sql/db/api.py
@@ -1522,6 +1522,31 @@ def add_test_run_attachments(attach_dict, test_run_id, session=None):
     return attachments
 
 
+def get_recent_failed_runs_by_run_metadata(key, value, num_runs=10,
+                                           start_date=None, session=None):
+    """Get a list of recent failed runs for a given run metadata pair
+
+    :param str key: the run_metadata key to get failed runs
+    :param str value: The run_metadata value to get failed runs
+    :param int num_runs: The number of results to fetch, defaults to 10
+    :param datetime start_date: the optional starting dates to get runs from.
+                                Nothing older than this date will be returned
+    :param session: optional session object if one isn't provided a new session
+                    will be acquired for the duration of this operation
+
+    :return list: The list of recent failed Run objects
+    :rtype: subunit2sql.models.Run
+    """
+    session = session or get_session()
+    query = db_utils.model_query(models.Run, session).join(
+        models.RunMetadata, models.Run.id == models.RunMetadata.run_id).filter(
+            models.RunMetadata.key == key,
+            models.RunMetadata.value == value)
+    query = _filter_runs_by_date(query, start_date)
+    return query.filter(models.Run.fails > 0).order_by(
+        models.Run.run_at.desc()).limit(num_runs).all()
+
+
 def get_runs_by_status_grouped_by_run_metadata(key, start_date=None,
                                                stop_date=None, session=None):
     session = session or get_session()
diff --git a/subunit2sql/tests/db/test_api.py b/subunit2sql/tests/db/test_api.py
index 766cc3f..92843a7 100644
--- a/subunit2sql/tests/db/test_api.py
+++ b/subunit2sql/tests/db/test_api.py
@@ -790,3 +790,28 @@ class TestDatabaseAPI(base.TestCase):
         for meta in test_metadata:
             self.assertIn(meta.key, test_meta.keys())
             self.assertIn(meta.value, test_meta.values())
+
+    def test_get_recent_failed_runs_by_run_metadata_no_start_date(self):
+        run_a = api.create_run(fails=1)
+        api.create_run()
+        run_c = api.create_run(fails=2)
+        api.add_run_metadata({'fake_key': 'fake_value'}, run_a.id)
+        api.add_run_metadata({'zeon': 'zaku'}, run_c.id)
+        results = api.get_recent_failed_runs_by_run_metadata('zeon', 'zaku')
+        self.assertEqual(1, len(results))
+        self.assertEqual(run_c.id, results[0].id)
+
+    def test_get_recent_failed_runs_by_run_metadata_with_start_date(self):
+        run_a = api.create_run(fails=1)
+        api.create_run()
+        run_c = api.create_run(fails=2)
+        run_d = api.create_run(fails=3,
+                               run_at=datetime.datetime(1914, 6, 28,
+                                                        10, 45, 0))
+        api.add_run_metadata({'fake_key': 'fake_value'}, run_a.id)
+        api.add_run_metadata({'zeon': 'zaku'}, run_c.id)
+        api.add_run_metadata({'zeon': 'zaku'}, run_d.id)
+        results = api.get_recent_failed_runs_by_run_metadata(
+            'zeon', 'zaku', start_date=datetime.date(1970, 1, 1))
+        self.assertEqual(1, len(results))
+        self.assertEqual(run_c.id, results[0].id)