diff --git a/xarray/core/combine.py b/xarray/core/combine.py
index 573247937b7..375931e1f9c 100644
--- a/xarray/core/combine.py
+++ b/xarray/core/combine.py
@@ -44,6 +44,18 @@ def _infer_tile_ids_from_nested_list(entry, current_pos):
         yield current_pos, entry
 
 
+def _ensure_same_types(series, dim):
+
+    if series.dtype == object:
+        types = set(series.map(type))
+        if len(types) > 1:
+            types = ", ".join(t.__name__ for t in types)
+            raise TypeError(
+                f"Cannot combine along dimension '{dim}' with mixed types."
+                f" Found: {types}."
+            )
+
+
 def _infer_concat_order_from_coords(datasets):
 
     concat_dims = []
@@ -88,11 +100,15 @@ def _infer_concat_order_from_coords(datasets):
                     raise ValueError("Cannot handle size zero dimensions")
                 first_items = pd.Index([index[0] for index in indexes])
 
+                series = first_items.to_series()
+
+                # ensure series does not contain mixed types, e.g. cftime calendars
+                _ensure_same_types(series, dim)
+
                 # Sort datasets along dim
                 # We want rank but with identical elements given identical
                 # position indices - they should be concatenated along another
                 # dimension, not along this one
-                series = first_items.to_series()
                 rank = series.rank(
                     method="dense", ascending=ascending, numeric_only=False
                 )
diff --git a/xarray/tests/test_combine.py b/xarray/tests/test_combine.py
index 522b98cf864..7664e5e56db 100644
--- a/xarray/tests/test_combine.py
+++ b/xarray/tests/test_combine.py
@@ -1,4 +1,5 @@
 from datetime import datetime
+from distutils.version import LooseVersion
 from itertools import product
 
 import numpy as np
@@ -865,5 +866,22 @@ def test_combine_by_coords_raises_for_differing_calendars():
     da_1 = DataArray([0], dims=["time"], coords=[time_1], name="a").to_dataset()
     da_2 = DataArray([1], dims=["time"], coords=[time_2], name="a").to_dataset()
 
-    with raises_regex(TypeError, r"cannot compare .* \(different calendars\)"):
+    if LooseVersion(cftime.__version__) >= LooseVersion("1.5"):
+        error_msg = "Cannot combine along dimension 'time' with mixed types."
+    else:
+        error_msg = r"cannot compare .* \(different calendars\)"
+
+    with raises_regex(TypeError, error_msg):
+        combine_by_coords([da_1, da_2])
+
+
+def test_combine_by_coords_raises_for_differing_types():
+
+    # str and byte cannot be compared
+    da_1 = DataArray([0], dims=["time"], coords=[["a"]], name="a").to_dataset()
+    da_2 = DataArray([1], dims=["time"], coords=[[b"b"]], name="a").to_dataset()
+
+    with raises_regex(
+        TypeError, "Cannot combine along dimension 'time' with mixed types."
+    ):
         combine_by_coords([da_1, da_2])