diff --git a/tests/pipeline/test_blaze.py b/tests/pipeline/test_blaze.py
index 105ed219..a0faf6be 100644
--- a/tests/pipeline/test_blaze.py
+++ b/tests/pipeline/test_blaze.py
@@ -30,7 +30,6 @@ from zipline.pipeline.loaders.blaze import (
     NoDeltasWarning,
 )
 from zipline.pipeline.loaders.blaze.core import (
-    NonNumpyField,
     NonPipelineField,
     no_deltas_rules,
 )
@@ -277,11 +276,11 @@ class BlazeToPipelineTestCase(TestCase):
         )
 
     def test_cols_with_missing_vals(self):
-        dates = (pd.Timestamp('2014-01-01'), pd.Timestamp('2014-01-03')) * 3
+        dates = (pd.Timestamp('2014-01-01'), pd.Timestamp('2014-01-03'))
         df = pd.DataFrame({
-            'sid': self.sids * 2,
-            'value': (np.NaN, 0., 1., 2., 3., 2.,),
-            'str_value': (None, "b", "c", "d", "e", "f"),
+            'sid': self.sids[:-1],
+            'value': (0., 1.,),
+            'str_value': ("a", "b",),
             'asof_date': dates,
             'timestamp': dates,
         })
@@ -299,7 +298,21 @@ class BlazeToPipelineTestCase(TestCase):
         fields = OrderedDict(expr.dshape.measure.fields)
 
         with tmp_asset_finder() as finder:
-            expected = pd.DataFrame()
+            expected = pd.DataFrame(
+                np.array([["a", 0],
+                          [None, np.NaN],
+                          [None, np.NaN],
+                          ["a", 0],
+                          [None, np.NaN],
+                          [None, np.NaN],
+                          ["a", 0],
+                          ["b", 1],
+                          [None, np.NaN]]),
+                columns=['str_value', 'value'],
+                index=pd.MultiIndex.from_product(
+                    (self.dates, finder.retrieve_all(self.sids))
+                )
+            )
             self._test_id(
                 df,
                 var * Record(fields),
@@ -369,7 +382,6 @@ class BlazeToPipelineTestCase(TestCase):
                 dates,
                 finder,
             ).run_pipeline(p, dates[0], dates[-1])
-        import pdb; pdb.set_trace()
         assert_frame_equal(
             result,
             _utc_localize_index_level_0(expected),
diff --git a/zipline/pipeline/loaders/blaze/core.py b/zipline/pipeline/loaders/blaze/core.py
index fd265691..53504fe7 100644
--- a/zipline/pipeline/loaders/blaze/core.py
+++ b/zipline/pipeline/loaders/blaze/core.py
@@ -177,7 +177,7 @@ from zipline.utils.input_validation import (
     ensure_timezone,
     optionally,
 )
-from zipline.utils.numpy_utils import repeat_last_axis, categorical_dtype
+from zipline.utils.numpy_utils import categorical_dtype, repeat_last_axis
 from zipline.utils.pandas_utils import sort_values
 from zipline.utils.preprocess import preprocess
 
@@ -314,7 +314,7 @@ def new_dataset(expr, deltas, missing_values):
             if isinstance(type_, Option):
                 type_ = type_.ty
             type_ = type_.to_numpy_dtype()
-            if not isinstance(type_, String) and not can_represent_dtype(type_):
+            if not can_represent_dtype(type_):
                 raise NotPipelineCompatible()
             col = Column(
                 type_,
@@ -1006,15 +1006,15 @@ class BlazeLoader(dict):
                     )
                 else:
                     last_in_group = last_in_group.reindex(dates)
-            import pdb; pdb.set_trace()
-            # str_cols = df.columns[df.dtypes == categorical_dtype]
-            #
-            # # Unstack will fill all missing values with NaN; we need to fix
-            # # this for strings.
-            # for col in str_cols:
-            #     last_in_group.iloc[
-            #         :, last_in_group.columns.get_level_values(0) == col
-            #     ].fillna('None')
+            # Unstack will fill all missing values with NaN; we need to fix
+            # this for strings.
+            if not df.empty:
+                str_cols = df.columns[df.dtypes == categorical_dtype]
+
+                for col in str_cols:
+                    last_in_group[col] = last_in_group[col].where(pd.notnull(
+                        last_in_group[col]), None)
+
             return last_in_group
 
         sparse_deltas = last_in_date_group(non_novel_deltas, reindex=False)