diff --git a/pandas/tests/arrays/test_datetimelike.py b/pandas/tests/arrays/test_datetimelike.py
index 159f52a4c7c25..c489aa5867632 100644
--- a/pandas/tests/arrays/test_datetimelike.py
+++ b/pandas/tests/arrays/test_datetimelike.py
@@ -3,7 +3,6 @@
 
 import numpy as np
 import pytest
-import pytz
 
 from pandas._libs import NaT, OutOfBoundsDatetime, Timestamp
 from pandas.compat.numpy import np_version_under1p18
@@ -269,18 +268,16 @@ def test_searchsorted(self):
             assert result == 10
 
     @pytest.mark.parametrize("box", [None, "index", "series"])
-    def test_searchsorted_castable_strings(self, arr1d, box):
+    def test_searchsorted_castable_strings(self, arr1d, box, request):
         if isinstance(arr1d, DatetimeArray):
             tz = arr1d.tz
-            if (
-                tz is not None
-                and tz is not pytz.UTC
-                and not isinstance(tz, pytz._FixedOffset)
-            ):
+            ts1, ts2 = arr1d[1:3]
+            if tz is not None and ts1.tz.tzname(ts1) != ts2.tz.tzname(ts2):
                 # If we have e.g. tzutc(), when we cast to string and parse
                 #  back we get pytz.UTC, and then consider them different timezones
                 #  so incorrectly raise.
-                pytest.xfail(reason="timezone comparisons inconsistent")
+                mark = pytest.mark.xfail(reason="timezone comparisons inconsistent")
+                request.node.add_marker(mark)
 
         arr = arr1d
         if box is None:
@@ -391,19 +388,17 @@ def test_setitem(self):
         expected[:2] = expected[-2:]
         tm.assert_numpy_array_equal(arr.asi8, expected)
 
-    def test_setitem_strs(self, arr1d):
+    def test_setitem_strs(self, arr1d, request):
         # Check that we parse strs in both scalar and listlike
         if isinstance(arr1d, DatetimeArray):
             tz = arr1d.tz
-            if (
-                tz is not None
-                and tz is not pytz.UTC
-                and not isinstance(tz, pytz._FixedOffset)
-            ):
+            ts1, ts2 = arr1d[-2:]
+            if tz is not None and ts1.tz.tzname(ts1) != ts2.tz.tzname(ts2):
                 # If we have e.g. tzutc(), when we cast to string and parse
                 #  back we get pytz.UTC, and then consider them different timezones
                 #  so incorrectly raise.
-                pytest.xfail(reason="timezone comparisons inconsistent")
+                mark = pytest.mark.xfail(reason="timezone comparisons inconsistent")
+                request.node.add_marker(mark)
 
         # Setting list-like of strs
         expected = arr1d.copy()
diff --git a/pandas/tests/base/test_conversion.py b/pandas/tests/base/test_conversion.py
index 668954a3f4a0b..cc4aed5e4413d 100644
--- a/pandas/tests/base/test_conversion.py
+++ b/pandas/tests/base/test_conversion.py
@@ -328,7 +328,7 @@ def test_array_multiindex_raises():
         ),
     ],
 )
-def test_to_numpy(array, expected, index_or_series_or_array):
+def test_to_numpy(array, expected, index_or_series_or_array, request):
     box = index_or_series_or_array
     thing = box(array)
 
@@ -336,7 +336,8 @@ def test_to_numpy(array, expected, index_or_series_or_array):
         pytest.skip(f"No index type for {array.dtype}")
 
     if array.dtype.name == "int64" and box is pd.array:
-        pytest.xfail("thing is Int64 and to_numpy() returns object")
+        mark = pytest.mark.xfail(reason="thing is Int64 and to_numpy() returns object")
+        request.node.add_marker(mark)
 
     result = thing.to_numpy()
     tm.assert_numpy_array_equal(result, expected)
diff --git a/pandas/tests/indexes/test_numpy_compat.py b/pandas/tests/indexes/test_numpy_compat.py
index b71417b2a625d..8bfb97ca494e6 100644
--- a/pandas/tests/indexes/test_numpy_compat.py
+++ b/pandas/tests/indexes/test_numpy_compat.py
@@ -74,14 +74,16 @@ def test_numpy_ufuncs_basic(index, func):
 @pytest.mark.parametrize(
     "func", [np.isfinite, np.isinf, np.isnan, np.signbit], ids=lambda x: x.__name__
 )
-def test_numpy_ufuncs_other(index, func):
+def test_numpy_ufuncs_other(index, func, request):
     # test ufuncs of numpy, see:
     # https://numpy.org/doc/stable/reference/ufuncs.html
 
     if isinstance(index, (DatetimeIndex, TimedeltaIndex)):
         if isinstance(index, DatetimeIndex) and index.tz is not None:
             if func in [np.isfinite, np.isnan, np.isinf]:
-                pytest.xfail(reason="__array_ufunc__ is not defined")
+                if not np_version_under1p17:
+                    mark = pytest.mark.xfail(reason="__array_ufunc__ is not defined")
+                    request.node.add_marker(mark)
 
         if not np_version_under1p18 and func in [np.isfinite, np.isinf, np.isnan]:
             # numpy 1.18(dev) changed isinf and isnan to not raise on dt64/tfd64
diff --git a/pandas/tests/series/test_arithmetic.py b/pandas/tests/series/test_arithmetic.py
index 09109447e799a..b204d92b9122f 100644
--- a/pandas/tests/series/test_arithmetic.py
+++ b/pandas/tests/series/test_arithmetic.py
@@ -756,12 +756,15 @@ def test_align_date_objects_with_datetimeindex(self):
 )
 @pytest.mark.parametrize("box", [list, tuple, np.array, pd.Index, pd.Series, pd.array])
 @pytest.mark.parametrize("flex", [True, False])
-def test_series_ops_name_retention(flex, box, names, all_binary_operators):
+def test_series_ops_name_retention(flex, box, names, all_binary_operators, request):
     # GH#33930 consistent name retention
     op = all_binary_operators
 
-    if op is ops.rfloordiv and box in [list, tuple]:
-        pytest.xfail("op fails because of inconsistent ndarray-wrapping GH#28759")
+    if op is ops.rfloordiv and box in [list, tuple] and not flex:
+        mark = pytest.mark.xfail(
+            reason="op fails because of inconsistent ndarray-wrapping GH#28759"
+        )
+        request.node.add_marker(mark)
 
     left = Series(range(10), name=names[0])
     right = Series(range(10), name=names[1])