diff --git a/django/db/models/base.py b/django/db/models/base.py
index d7a9932388..fd7250cdb0 100644
--- a/django/db/models/base.py
+++ b/django/db/models/base.py
@@ -404,8 +404,6 @@ class Model(six.with_metaclass(ModelBase, object)):
             u = six.text_type(self)
         except (UnicodeEncodeError, UnicodeDecodeError):
             u = '[Bad Unicode data]'
-        if not six.PY3:
-            u = u.encode('ascii', 'replace')
         return smart_str('<%s: %s>' % (self.__class__.__name__, u))
 
     def __str__(self):
diff --git a/tests/regressiontests/model_regress/tests.py b/tests/regressiontests/model_regress/tests.py
index 6a03b861e4..6a45a83052 100644
--- a/tests/regressiontests/model_regress/tests.py
+++ b/tests/regressiontests/model_regress/tests.py
@@ -1,5 +1,3 @@
-# coding: utf-8
-
 from __future__ import absolute_import, unicode_literals
 
 import datetime
@@ -148,14 +146,6 @@ class ModelTests(TestCase):
         b = BrokenUnicodeMethod.objects.create(name="Jerry")
         self.assertEqual(repr(b), "<BrokenUnicodeMethod: [Bad Unicode data]>")
 
-    def test_no_unicode_in_repr(self):
-        a = Article.objects.create(
-            headline="Watch for umlauts: üöä", pub_date=datetime.datetime.now())
-        if six.PY3:
-            self.assertEqual(repr(a), '<Article: Watch for umlauts: üöä>')
-        else:
-            self.assertEqual(repr(a), '<Article: Watch for umlauts: ???>')
-
     @skipUnlessDBFeature("supports_timezones")
     def test_timezones(self):
         # Saving an updating with timezone-aware datetime Python objects.