diff --git a/moto/s3/responses.py b/moto/s3/responses.py
index ee047a14..4c546c59 100644
--- a/moto/s3/responses.py
+++ b/moto/s3/responses.py
@@ -451,17 +451,16 @@ class ResponseObject(_TemplateEnvironmentMixin, ActionAuthenticatorMixin):
         continuation_token = querystring.get('continuation-token', [None])[0]
         start_after = querystring.get('start-after', [None])[0]
 
+        # sort the combination of folders and keys into lexicographical order
+        all_keys = result_keys + result_folders
+        all_keys.sort(key=self._get_name)
+
         if continuation_token or start_after:
             limit = continuation_token or start_after
-            if not delimiter:
-                result_keys = self._get_results_from_token(result_keys, limit)
-            else:
-                result_folders = self._get_results_from_token(result_folders, limit)
+            all_keys = self._get_results_from_token(all_keys, limit)
 
-        if not delimiter:
-            result_keys, is_truncated, next_continuation_token = self._truncate_result(result_keys, max_keys)
-        else:
-            result_folders, is_truncated, next_continuation_token = self._truncate_result(result_folders, max_keys)
+        truncated_keys, is_truncated, next_continuation_token = self._truncate_result(all_keys, max_keys)
+        result_keys, result_folders = self._split_truncated_keys(truncated_keys)
 
         key_count = len(result_keys) + len(result_folders)
 
@@ -479,6 +478,24 @@ class ResponseObject(_TemplateEnvironmentMixin, ActionAuthenticatorMixin):
             start_after=None if continuation_token else start_after
         )
 
+    @staticmethod
+    def _get_name(key):
+        if isinstance(key, FakeKey):
+            return key.name
+        else:
+            return key
+
+    @staticmethod
+    def _split_truncated_keys(truncated_keys):
+        result_keys = []
+        result_folders = []
+        for key in truncated_keys:
+            if isinstance(key, FakeKey):
+                result_keys.append(key)
+            else:
+                result_folders.append(key)
+        return result_keys, result_folders
+
     def _get_results_from_token(self, result_keys, token):
         continuation_index = 0
         for key in result_keys:
diff --git a/tests/test_s3/test_s3.py b/tests/test_s3/test_s3.py
index 0c0721f0..2764ee2c 100644
--- a/tests/test_s3/test_s3.py
+++ b/tests/test_s3/test_s3.py
@@ -1392,6 +1392,34 @@ def test_boto3_list_objects_v2_fetch_owner():
     assert len(owner.keys()) == 2
 
 
+@mock_s3
+def test_boto3_list_objects_v2_truncate_combined_keys_and_folders():
+    s3 = boto3.client('s3', region_name='us-east-1')
+    s3.create_bucket(Bucket='mybucket')
+    s3.put_object(Bucket='mybucket', Key='1/2', Body='')
+    s3.put_object(Bucket='mybucket', Key='2', Body='')
+    s3.put_object(Bucket='mybucket', Key='3/4', Body='')
+    s3.put_object(Bucket='mybucket', Key='4', Body='')
+
+    resp = s3.list_objects_v2(Bucket='mybucket', Prefix='', MaxKeys=2, Delimiter='/')
+    assert 'Delimiter' in resp
+    assert resp['IsTruncated'] is True
+    assert resp['KeyCount'] == 2
+    assert len(resp['Contents']) == 1
+    assert resp['Contents'][0]['Key'] == '2'
+    assert len(resp['CommonPrefixes']) == 1
+    assert resp['CommonPrefixes'][0]['Prefix'] == '1/'
+
+    last_tail = resp['NextContinuationToken']
+    resp = s3.list_objects_v2(Bucket='mybucket', MaxKeys=2, Prefix='', Delimiter='/', StartAfter=last_tail)
+    assert resp['KeyCount'] == 2
+    assert resp['IsTruncated'] is False
+    assert len(resp['Contents']) == 1
+    assert resp['Contents'][0]['Key'] == '4'
+    assert len(resp['CommonPrefixes']) == 1
+    assert resp['CommonPrefixes'][0]['Prefix'] == '3/'
+
+
 @mock_s3
 def test_boto3_bucket_create():
     s3 = boto3.resource('s3', region_name='us-east-1')