diff --git a/pyuvdata/uvcal/tests/test_calh5.py b/pyuvdata/uvcal/tests/test_calh5.py
index a17a24f40d..1acf234346 100644
--- a/pyuvdata/uvcal/tests/test_calh5.py
+++ b/pyuvdata/uvcal/tests/test_calh5.py
@@ -11,9 +11,11 @@
 
 import pyuvdata.tests as uvtest
 from pyuvdata import UVCal
+from pyuvdata.data import DATA_PATH
 from pyuvdata.uvcal import FastCalH5Meta
 from pyuvdata.uvcal.tests import extend_jones_axis, time_array_to_time_range
 from pyuvdata.uvcal.uvcal import _future_array_shapes_warning
+from pyuvdata.uvdata import FastUVH5Meta
 
 
 @pytest.mark.filterwarnings("ignore:" + _future_array_shapes_warning)
@@ -114,6 +116,18 @@ def test_calh5_meta(gain_data, tmp_path):
     calobj.write_calh5(write_file, clobber=True)
     cal_meta = FastCalH5Meta(write_file)
 
+    cal_meta2 = FastCalH5Meta(write_file)
+    assert cal_meta == cal_meta2
+
+    write_file2 = str(tmp_path / "outtest2.calh5")
+    calobj.write_calh5(write_file, clobber=True)
+
+    cal_meta2 = FastCalH5Meta(write_file2)
+    assert cal_meta != cal_meta2
+
+    uvh5_meta = FastUVH5Meta(os.path.join(DATA_PATH, "zen.2458432.34569.uvh5"))
+    assert cal_meta != uvh5_meta
+
     ant_nums = cal_meta.antenna_numbers
     jpol_nums = cal_meta.jones_array
     jpol_names = cal_meta.pols
diff --git a/pyuvdata/uvdata/uvh5.py b/pyuvdata/uvdata/uvh5.py
index a8fd89ad6e..86f99acb70 100644
--- a/pyuvdata/uvdata/uvh5.py
+++ b/pyuvdata/uvdata/uvh5.py
@@ -394,27 +394,6 @@ def channel_width(self) -> float:
         else:
             return float(h["channel_width"][()])
 
-    @cached_property
-    def extra_keywords(self) -> dict:
-        """The extra_keywords from the file."""
-        header = self.header
-        if "extra_keywords" not in header:
-            return {}
-
-        extra_keywords = {}
-        for key in header["extra_keywords"].keys():
-            if header["extra_keywords"][key].dtype.type in (np.string_, np.object_):
-                extra_keywords[key] = bytes(header["extra_keywords"][key][()]).decode(
-                    "utf8"
-                )
-            else:
-                # special handling for empty datasets == python `None` type
-                if header["extra_keywords"][key].shape is None:
-                    extra_keywords[key] = None
-                else:
-                    extra_keywords[key] = header["extra_keywords"][key][()]
-        return extra_keywords
-
     @cached_property
     def unique_antpair_1_array(self) -> np.ndarray:
         """The unique antenna 1 indices in the file."""