diff --git a/doc/data-structures.rst b/doc/data-structures.rst
index 74617a43340..947c09db7d6 100644
--- a/doc/data-structures.rst
+++ b/doc/data-structures.rst
@@ -368,9 +368,10 @@ Transforming datasets
 In addition to dictionary-like methods (described above), xarray has additional
 methods (like pandas) for transforming datasets into new objects.
 
-For removing variables, you can select and drop an explicit list of variables
-by indexing with a list of names or using the :py:meth:`~xarray.Dataset.drop`
-methods to return a new ``Dataset``. These operations keep around coordinates:
+For removing variables, you can select and drop an explicit list of
+variables by indexing with a list of names or using the
+:py:meth:`~xray.Dataset.drop` methods to return a new ``Dataset``. These
+operations keep around coordinates:
 
 .. ipython:: python
 
diff --git a/doc/whats-new.rst b/doc/whats-new.rst
index a6567193951..9c6c6a6089b 100644
--- a/doc/whats-new.rst
+++ b/doc/whats-new.rst
@@ -75,6 +75,8 @@ By `Robin Wilson <https://github.com/robintw>`_.
 
 Bug fixes
 ~~~~~~~~~
+- ``groupby_bins`` now restores empty bins by default (:issue:`1019`).
+  By `Ryan Abernathey <https://github.com/rabernat>`_.
 
 - Fix issues for dates outside the valid range of pandas timestamps
   (:issue:`975`). By `Mathias Hauser <https://github.com/mathause>`_.
diff --git a/xarray/core/groupby.py b/xarray/core/groupby.py
index 7b925af2105..2c03cf5421f 100644
--- a/xarray/core/groupby.py
+++ b/xarray/core/groupby.py
@@ -154,6 +154,7 @@ def __init__(self, obj, group, squeeze=False, grouper=None, bins=None,
             specified bins by `pandas.cut`.
         cut_kwargs : dict, optional
             Extra keyword arguments to pass to `pandas.cut`
+
         """
         from .dataset import as_dataset
         from .dataarray import DataArray
@@ -193,6 +194,7 @@ def __init__(self, obj, group, squeeze=False, grouper=None, bins=None,
             binned = pd.cut(group.values, bins, **cut_kwargs)
             new_dim_name = group.name + '_bins'
             group = DataArray(binned, group.coords, name=new_dim_name)
+            full_index = binned.categories
         if grouper is not None:
             index = safe_cast_to_index(group)
             if not index.is_monotonic:
diff --git a/xarray/test/test_dataarray.py b/xarray/test/test_dataarray.py
index f613259f128..79b5c893039 100644
--- a/xarray/test/test_dataarray.py
+++ b/xarray/test/test_dataarray.py
@@ -1416,6 +1416,18 @@ def test_groupby_bins(self):
         # (would fail with shortcut=True above)
         self.assertEqual(len(array.dim_0), 4)
 
+    def test_groupby_bins_empty(self):
+        array = DataArray(np.arange(4), dims='dim_0')
+        # one of these bins will be empty
+        bins = [0,4,5]
+        actual = array.groupby_bins('dim_0', bins).sum()
+        expected = DataArray([6, np.nan], dims='dim_0_bins',
+                        coords={'dim_0_bins': ['(0, 4]','(4, 5]']})
+        self.assertDataArrayIdentical(expected, actual)
+        # make sure original array is unchanged
+        # (was a problem in earlier versions)
+        self.assertEqual(len(array.dim_0), 4)
+
     def test_groupby_bins_multidim(self):
         array = self.make_groupby_multidim_example_array()
         bins = [0,15,20]