diff --git a/doc/api.rst b/doc/api.rst
index b843af17900..8fe61c643ba 100644
--- a/doc/api.rst
+++ b/doc/api.rst
@@ -77,6 +77,8 @@ Computations
 .. autosummary::
    :toctree: generated/
 
+   Dataset.apply
+   Dataset.reduce
    Dataset.all
    Dataset.any
    Dataset.argmax
diff --git a/test/test_dataset.py b/test/test_dataset.py
index 6136e752027..992ffcb7be6 100644
--- a/test/test_dataset.py
+++ b/test/test_dataset.py
@@ -779,3 +779,32 @@ def test_reduce_keep_attrs(self):
         ds = data.mean(keep_attrs=True)
         self.assertEqual(len(ds.attrs), len(_attrs))
         self.assertTrue(ds.attrs, attrs)
+
+    def test_apply(self):
+        data = create_test_data()
+        data.attrs['foo'] = 'bar'
+
+        self.assertDatasetIdentical(data.apply(np.mean), data.mean())
+        self.assertDatasetIdentical(data.apply(np.mean, keep_attrs=True),
+                                    data.mean(keep_attrs=True))
+
+        self.assertDatasetIdentical(data.apply(lambda x: x, keep_attrs=True),
+                                    data.drop_vars('time'))
+
+        actual = data.apply(np.mean, to=['var1', 'var2', 'var3'])
+        self.assertDatasetIdentical(actual, data.mean())
+
+        actual = data.apply(np.mean, to='var1')
+        modified = data.select_vars('var1').mean()
+        unmodified = data.select_vars('var2', 'var3')
+        expected = modified.merge(unmodified)
+        self.assertDatasetIdentical(actual, expected)
+
+        with self.assertRaisesRegexp(ValueError, 'does not contain'):
+            data.apply(np.mean, to='foobarbaz')
+
+        def scale(x, multiple=1):
+            return multiple * x
+
+        actual = data.apply(scale, multiple=2)
+        self.assertDataArrayEqual(actual['var1'], 2 * data['var1'])
diff --git a/xray/dataset.py b/xray/dataset.py
index 6f0cfa3cb0f..843850816fb 100644
--- a/xray/dataset.py
+++ b/xray/dataset.py
@@ -1119,6 +1119,50 @@ def reduce(self, func, dimension=None, keep_attrs=False, **kwargs):
 
         return Dataset(variables=variables, attributes=attrs)
 
+    def apply(self, func, to=None, keep_attrs=False, **kwargs):
+        """Apply a function over noncoordinates in this dataset.
+
+        Parameters
+        ----------
+        func : function
+            Function which can be called in the form `f(x, **kwargs)` to
+            transform each DataArray `x` in this dataset into another
+            DataArray.
+        to : str or sequence of str, optional
+            Explicit list of noncoordinates in this dataset to which to apply
+            `func`. Unlisted noncoordinates are passed through unchanged. By
+            default, `func` is applied to all noncoordinates in this dataset.
+        keep_attrs : bool, optional
+            If True, the datasets's attributes (`attrs`) will be copied from
+            the original object to the new one. If False, the new object will
+            be returned without attributes.
+        **kwargs : dict
+            Additional keyword arguments passed on to `func`.
+
+        Returns
+        -------
+        applied : Dataset
+            Resulting dataset from applying over each noncoordinate.
+            Coordinates which are no longer used as the dimension of a
+            noncoordinate are dropped.
+        """
+        if to is not None:
+            to = set([to] if isinstance(to, basestring) else to)
+            bad_to = to - set(self.noncoordinates)
+            if bad_to:
+                raise ValueError('Dataset does not contain the '
+                                 'noncoordinates: %r' % list(bad_to))
+        else:
+            to = set(self.noncoordinates)
+
+        variables = OrderedDict()
+        for name, var in iteritems(self.noncoordinates):
+            variables[name] = func(var, **kwargs) if name in to else var
+
+        attrs = self.attrs if keep_attrs else {}
+
+        return Dataset(variables, attrs)
+
     @classmethod
     def concat(cls, datasets, dimension='concat_dimension', indexers=None,
                mode='different', concat_over=None, compat='equals'):