diff --git a/AUTHORS b/AUTHORS
index 05be5d870c..bcf23acb16 100644
--- a/AUTHORS
+++ b/AUTHORS
@@ -620,6 +620,7 @@ answer newbie questions, and generally made Django that much better:
     Maximillian Dornseif <md@hudora.de>
     mccutchen@gmail.com
     Meir Kriheli <http://mksoft.co.il/>
+    Michael S. Brown <michael@msbrown.net>
     Michael Hall <mhall1@ualberta.ca>
     Michael Josephson <http://www.sdjournal.com/>
     Michael Manfre <mmanfre@gmail.com>
diff --git a/django/http/multipartparser.py b/django/http/multipartparser.py
index 33d8814241..b3472f7be2 100644
--- a/django/http/multipartparser.py
+++ b/django/http/multipartparser.py
@@ -9,6 +9,7 @@ import binascii
 import cgi
 import collections
 import html
+import os
 from urllib.parse import unquote
 
 from django.conf import settings
@@ -208,6 +209,7 @@ class MultiPartParser:
                     # This is a file, use the handler...
                     file_name = disposition.get('filename')
                     if file_name:
+                        file_name = os.path.basename(file_name)
                         file_name = force_str(file_name, encoding, errors='replace')
                         file_name = self.IE_sanitize(html.unescape(file_name))
                     if not file_name:
diff --git a/tests/file_uploads/tests.py b/tests/file_uploads/tests.py
index 5743344a51..df8f1b8031 100644
--- a/tests/file_uploads/tests.py
+++ b/tests/file_uploads/tests.py
@@ -209,10 +209,14 @@ class FileUploadTests(TestCase):
         Receiving file upload when filename is blank (before and after
         sanitization) should be okay.
         """
-        # The second value is normalized to an empty name by
-        # MultiPartParser.IE_sanitize()
-        filenames = ['', 'C:\\Windows\\']
-
+        filenames = [
+            '',
+            # Normalized by MultiPartParser.IE_sanitize().
+            'C:\\Windows\\',
+            # Normalized by os.path.basename().
+            '/',
+            'ends-with-slash/',
+        ]
         payload = client.FakePayload()
         for i, name in enumerate(filenames):
             payload.write('\r\n'.join([