diff --git a/mapping/tests/test_util.py b/mapping/tests/test_util.py
index cc2a57e..76dce60 100644
--- a/mapping/tests/test_util.py
+++ b/mapping/tests/test_util.py
@@ -766,3 +766,11 @@ def test_reindex(self):
 
         exp_rets = pd.Series([0.02, np.NaN], index=widx)
         assert_series_equal(exp_rets, new_rets)
+
+        # check unique index to avoid duplicates from pd.Series.reindex
+        idx = pd.MultiIndex.from_tuples([(TS('2015-01-03'), 'CLF5'),
+                                         (TS('2015-01-04'), 'CLF5')])
+        returns = pd.Series([0.02, -0.02], index=idx)
+        widx = pd.MultiIndex.from_tuples([(TS('2015-01-03'), 'CLF5'),
+                                          (TS('2015-01-03'), 'CLF5')])
+        self.assertRaises(ValueError, util.reindex, returns, widx, limit=1)
diff --git a/mapping/util.py b/mapping/util.py
index ad0ee73..fbe63df 100644
--- a/mapping/util.py
+++ b/mapping/util.py
@@ -326,6 +326,9 @@ def reindex(returns, index, limit):
     to be dropped since it is not present in weights. A discussion of this
     issue available at https://github.com/matthewgilbert/mapping/issues/9
     """
+    if not index.is_unique:
+        raise ValueError("'index' must be unique")
+
     cumulative_rets = (returns + 1).groupby(level=1).cumprod()
     # reindexing can both drop days and introduce NaNs for days not present
     cumulative_rets = cumulative_rets.reindex(index)