diff --git a/numpy/core/_internal.py b/numpy/core/_internal.py
index 92ab0c8b08ff..fbe580deeab8 100644
--- a/numpy/core/_internal.py
+++ b/numpy/core/_internal.py
@@ -295,7 +295,11 @@ def _index_fields(ary, fields):
     view_dtype = {'names':names, 'formats':formats, 'offsets':offsets, 'itemsize':dt.itemsize}
     view = ary.view(dtype=view_dtype)
 
-    return view.copy()
+    # Return a copy for now until behavior is fully deprecated
+    # in favor of returning view
+    copy_dtype = {'names':view_dtype['names'], 'formats':view_dtype['formats']}
+    from numpy import array
+    return array(view, dtype=copy_dtype, copy=True)
 
 # Given a string containing a PEP 3118 format specifier,
 # construct a Numpy dtype
diff --git a/numpy/core/tests/test_multiarray.py b/numpy/core/tests/test_multiarray.py
index b9fd3ad863fc..118f221ae64d 100644
--- a/numpy/core/tests/test_multiarray.py
+++ b/numpy/core/tests/test_multiarray.py
@@ -1956,6 +1956,11 @@ def test_field_names(self):
             assert_equal(b[['f1','f2']][0].tolist(), (2, 3))
             assert_equal(b[['f2','f1']][0].tolist(), (3, 2))
             assert_equal(b[['f1','f3']][0].tolist(), (2, (1,)))
+            # view of subfield view/copy
+            assert_equal(b[['f1','f2']][0].view(('i4',2)).tolist(), (2, 3))
+            assert_equal(b[['f2','f1']][0].view(('i4',2)).tolist(), (3, 2))
+            view_dtype=[('f1', 'i4'),('f3', [('', 'i4')])]
+            assert_equal(b[['f1','f3']][0].view(view_dtype).tolist(), (2, (1,)))
         # non-ascii unicode field indexing is well behaved
         if not is_py3:
             raise SkipTest('non ascii unicode field indexing skipped; '