diff --git a/doc/whats-new.rst b/doc/whats-new.rst
index 20f9fd7ca2f..828712a8d13 100644
--- a/doc/whats-new.rst
+++ b/doc/whats-new.rst
@@ -15,6 +15,19 @@ What's New
 
 .. _whats-new.0.10.0:
 
+v0.10.0 (unreleased)
+--------------------
+
+Bug fixes
+~~~~~~~~~
+
+- Fixed unexpected behavior in ``Dataset.set_index()`` and
+  ``DataArray.set_index()`` introduced by Pandas 0.21.0. Setting a new
+  index with a single variable resulted in 1-level
+  ``pandas.MultiIndex`` instead of a simple ``pandas.Index``
+  (:issue:`1722`).  By `Benoit Bovy <https://github.com/benbovy>`_.
+
+
 v0.10.0 rc2 (13 November 2017)
 ------------------------------
 
diff --git a/xarray/core/dataset.py b/xarray/core/dataset.py
index 56c9df0af93..9c99213dc23 100644
--- a/xarray/core/dataset.py
+++ b/xarray/core/dataset.py
@@ -136,6 +136,14 @@ def merge_indexes(
         names, labels, levels = [], [], []
         current_index_variable = variables.get(dim)
 
+        for n in var_names:
+            var = variables[n]
+            if (current_index_variable is not None and
+                    var.dims != current_index_variable.dims):
+                raise ValueError(
+                    "dimension mismatch between %r %s and %r %s"
+                    % (dim, current_index_variable.dims, n, var.dims))
+
         if current_index_variable is not None and append:
             current_index = current_index_variable.to_index()
             if isinstance(current_index, pd.MultiIndex):
@@ -148,20 +156,19 @@ def merge_indexes(
                 labels.append(cat.codes)
                 levels.append(cat.categories)
 
-        for n in var_names:
-            names.append(n)
-            var = variables[n]
-            if ((current_index_variable is not None) and
-                    (var.dims != current_index_variable.dims)):
-                raise ValueError(
-                    "dimension mismatch between %r %s and %r %s"
-                    % (dim, current_index_variable.dims, n, var.dims))
-            else:
+        if not len(names) and len(var_names) == 1:
+            idx = pd.Index(variables[var_names[0]].values)
+
+        else:
+            for n in var_names:
+                names.append(n)
+                var = variables[n]
                 cat = pd.Categorical(var.values, ordered=True)
                 labels.append(cat.codes)
                 levels.append(cat.categories)
 
-        idx = pd.MultiIndex(labels=labels, levels=levels, names=names)
+            idx = pd.MultiIndex(labels=labels, levels=levels, names=names)
+
         vars_to_replace[dim] = IndexVariable(dim, idx)
         vars_to_remove.extend(var_names)
 
diff --git a/xarray/tests/test_dataset.py b/xarray/tests/test_dataset.py
index 0ffdeb61419..192624b64f3 100644
--- a/xarray/tests/test_dataset.py
+++ b/xarray/tests/test_dataset.py
@@ -2007,6 +2007,12 @@ def test_set_index(self):
         ds.set_index(x=mindex.names, inplace=True)
         self.assertDatasetIdentical(ds, expected)
 
+        # ensure set_index with no existing index and a single data var given
+        # doesn't return multi-index
+        ds = Dataset(data_vars={'x_var': ('x', [0, 1, 2])})
+        expected = Dataset(coords={'x': [0, 1, 2]})
+        self.assertDataArrayIdentical(ds.set_index(x='x_var'), expected)
+
     def test_reset_index(self):
         ds = create_test_multiindex()
         mindex = ds['x'].to_index()