diff --git a/pandas/tests/arithmetic/conftest.py b/pandas/tests/arithmetic/conftest.py
index 2714b68fa6ff4..44e6cc664de6d 100644
--- a/pandas/tests/arithmetic/conftest.py
+++ b/pandas/tests/arithmetic/conftest.py
@@ -154,7 +154,7 @@ def box(request):
 @pytest.fixture(params=[pd.Index,
                         pd.Series,
                         pytest.param(pd.DataFrame,
-                                     marks=pytest.mark.xfail(strict=True))],
+                                     marks=pytest.mark.xfail)],
                 ids=id_func)
 def box_df_fail(request):
     """
@@ -167,7 +167,7 @@ def box_df_fail(request):
                         (pd.Series, False),
                         (pd.DataFrame, False),
                         pytest.param((pd.DataFrame, True),
-                                     marks=pytest.mark.xfail(strict=True))],
+                                     marks=pytest.mark.xfail)],
                 ids=id_func)
 def box_transpose_fail(request):
     """
diff --git a/pandas/tests/arithmetic/test_numeric.py b/pandas/tests/arithmetic/test_numeric.py
index c3cd9f0f43559..2d26959c65632 100644
--- a/pandas/tests/arithmetic/test_numeric.py
+++ b/pandas/tests/arithmetic/test_numeric.py
@@ -407,7 +407,7 @@ class TestMultiplicationDivision(object):
         pytest.param(pd.Index,
                      marks=pytest.mark.xfail(reason="Index.__div__ always "
                                                     "raises",
-                                             raises=TypeError, strict=True)),
+                                             raises=TypeError)),
         pd.Series,
         pd.DataFrame
     ], ids=lambda x: x.__name__)
diff --git a/pandas/tests/arithmetic/test_object.py b/pandas/tests/arithmetic/test_object.py
index 511d74a2e790c..da5055cc7b737 100644
--- a/pandas/tests/arithmetic/test_object.py
+++ b/pandas/tests/arithmetic/test_object.py
@@ -92,7 +92,7 @@ def test_add_extension_scalar(self, other, box, op):
     @pytest.mark.parametrize('box', [
         pytest.param(pd.Index,
                      marks=pytest.mark.xfail(reason="Does not mask nulls",
-                                             strict=True, raises=TypeError)),
+                                             raises=TypeError)),
         pd.Series,
         pd.DataFrame
     ], ids=lambda x: x.__name__)
@@ -109,7 +109,7 @@ def test_objarr_add_str(self, box):
     @pytest.mark.parametrize('box', [
         pytest.param(pd.Index,
                      marks=pytest.mark.xfail(reason="Does not mask nulls",
-                                             strict=True, raises=TypeError)),
+                                             raises=TypeError)),
         pd.Series,
         pd.DataFrame
     ], ids=lambda x: x.__name__)
diff --git a/pandas/tests/arrays/categorical/test_constructors.py b/pandas/tests/arrays/categorical/test_constructors.py
index a473f44d5d4aa..f1475c5022a86 100644
--- a/pandas/tests/arrays/categorical/test_constructors.py
+++ b/pandas/tests/arrays/categorical/test_constructors.py
@@ -538,8 +538,7 @@ def test_construction_with_ordered(self):
         cat = Categorical([0, 1, 2], ordered=True)
         assert cat.ordered
 
-    @pytest.mark.xfail(reason="Imaginary values not supported in Categorical",
-                       strict=True)
+    @pytest.mark.xfail(reason="Imaginary values not supported in Categorical")
     def test_constructor_imaginary(self):
         values = [1, 2, 3 + 1j]
         c1 = Categorical(values)
diff --git a/pandas/tests/arrays/sparse/test_array.py b/pandas/tests/arrays/sparse/test_array.py
index 619cd05128ddb..b8cef92f6a6d4 100644
--- a/pandas/tests/arrays/sparse/test_array.py
+++ b/pandas/tests/arrays/sparse/test_array.py
@@ -488,7 +488,7 @@ def test_astype_all(self, any_real_dtype):
             SparseArray(np.array([0, 1], dtype='datetime64[ns]'),
                         dtype=SparseDtype('datetime64[ns]',
                                           pd.Timestamp('1970'))),
-            marks=[pytest.mark.xfail(reason="NumPy-7619", strict=True)],
+            marks=[pytest.mark.xfail(reason="NumPy-7619")],
         ),
         (SparseArray([0, 1, 10]), str,
          SparseArray(['0', '1', '10'], dtype=SparseDtype(str, '0'))),
diff --git a/pandas/tests/extension/base/ops.py b/pandas/tests/extension/base/ops.py
index 2161214190be0..cd5e55d9871b2 100644
--- a/pandas/tests/extension/base/ops.py
+++ b/pandas/tests/extension/base/ops.py
@@ -72,8 +72,7 @@ def test_arith_series_with_scalar(self, data, all_arithmetic_operators):
         s = pd.Series(data)
         self.check_opname(s, op_name, s.iloc[0], exc=self.series_scalar_exc)
 
-    @pytest.mark.xfail(run=False, reason="_reduce needs implementation",
-                       strict=True)
+    @pytest.mark.xfail(run=False, reason="_reduce needs implementation")
     def test_arith_frame_with_scalar(self, data, all_arithmetic_operators):
         # frame & scalar
         op_name = all_arithmetic_operators
diff --git a/pandas/tests/extension/base/setitem.py b/pandas/tests/extension/base/setitem.py
index 479ab71676315..5c767c28643c9 100644
--- a/pandas/tests/extension/base/setitem.py
+++ b/pandas/tests/extension/base/setitem.py
@@ -166,8 +166,7 @@ def test_setitem_frame_invalid_length(self, data):
         with pytest.raises(ValueError, match=xpr):
             df['B'] = data[:5]
 
-    @pytest.mark.xfail(reason="GH#20441: setitem on extension types.",
-                       strict=True)
+    @pytest.mark.xfail(reason="GH#20441: setitem on extension types.")
     def test_setitem_tuple_index(self, data):
         s = pd.Series(data[:2], index=[(0, 0), (0, 1)])
         expected = pd.Series(data.take([1, 1]), index=s.index)
diff --git a/pandas/tests/extension/json/test_json.py b/pandas/tests/extension/json/test_json.py
index a9fb22bb72497..66e5f6b6dc732 100644
--- a/pandas/tests/extension/json/test_json.py
+++ b/pandas/tests/extension/json/test_json.py
@@ -148,7 +148,7 @@ def test_stack(self):
         rows since we consider `{}` NA, but `.astype(object)` doesn't.
         """
 
-    @pytest.mark.xfail(reason="dict for NA", strict=True)
+    @pytest.mark.xfail(reason="dict for NA")
     def test_unstack(self, data, index):
         # The base test has NaN for the expected NA value.
         # this matches otherwise
diff --git a/pandas/tests/extension/test_categorical.py b/pandas/tests/extension/test_categorical.py
index 279bfb5dc8eab..5b873b337880e 100644
--- a/pandas/tests/extension/test_categorical.py
+++ b/pandas/tests/extension/test_categorical.py
@@ -72,7 +72,7 @@ class TestDtype(base.BaseDtypeTests):
 
 
 class TestInterface(base.BaseInterfaceTests):
-    @pytest.mark.skip(reason="Memory usage doesn't match", strict=True)
+    @pytest.mark.skip(reason="Memory usage doesn't match")
     def test_memory_usage(self, data):
         # Is this deliberate?
         super(TestInterface, self).test_memory_usage(data)
@@ -87,9 +87,9 @@ class TestReshaping(base.BaseReshapingTests):
 
 
 class TestGetitem(base.BaseGetitemTests):
-    skip_take = pytest.mark.skip(reason="GH-20664.", strict=True)
+    skip_take = pytest.mark.skip(reason="GH-20664.")
 
-    @pytest.mark.skip(reason="Backwards compatibility", strict=True)
+    @pytest.mark.skip(reason="Backwards compatibility")
     def test_getitem_scalar(self, data):
         # CategoricalDtype.type isn't "correct" since it should
         # be a parent of the elements (object). But don't want
@@ -117,7 +117,7 @@ def test_take_non_na_fill_value(self, data_missing):
     def test_take_out_of_bounds_raises(self, data, allow_fill):
         return super().test_take_out_of_bounds_raises(data, allow_fill)
 
-    @pytest.mark.skip(reason="GH-20747. Unobserved categories.", strict=True)
+    @pytest.mark.skip(reason="GH-20747. Unobserved categories.")
     def test_take_series(self, data):
         super().test_take_series(data)
 
@@ -125,12 +125,11 @@ def test_take_series(self, data):
     def test_reindex_non_na_fill_value(self, data_missing):
         super().test_reindex_non_na_fill_value(data_missing)
 
-    @pytest.mark.skip(reason="Categorical.take buggy", strict=True)
+    @pytest.mark.skip(reason="Categorical.take buggy")
     def test_take_empty(self, data, na_value, na_cmp):
         super().test_take_empty(data, na_value, na_cmp)
 
-    @pytest.mark.skip(reason="test not written correctly for categorical",
-                      strict=True)
+    @pytest.mark.skip(reason="test not written correctly for categorical")
     def test_reindex(self, data, na_value):
         super().test_reindex(data, na_value)
 
@@ -141,11 +140,11 @@ class TestSetitem(base.BaseSetitemTests):
 
 class TestMissing(base.BaseMissingTests):
 
-    @pytest.mark.skip(reason="Not implemented", strict=True)
+    @pytest.mark.skip(reason="Not implemented")
     def test_fillna_limit_pad(self, data_missing):
         super().test_fillna_limit_pad(data_missing)
 
-    @pytest.mark.skip(reason="Not implemented", strict=True)
+    @pytest.mark.skip(reason="Not implemented")
     def test_fillna_limit_backfill(self, data_missing):
         super().test_fillna_limit_backfill(data_missing)
 
@@ -155,7 +154,7 @@ class TestReduce(base.BaseNoReduceTests):
 
 
 class TestMethods(base.BaseMethodsTests):
-    @pytest.mark.skip(reason="Unobserved categories included", strict=True)
+    @pytest.mark.skip(reason="Unobserved categories included")
     def test_value_counts(self, all_data, dropna):
         return super().test_value_counts(all_data, dropna)
 
@@ -175,7 +174,7 @@ def test_combine_add(self, data_repeated):
         expected = pd.Series([a + val for a in list(orig_data1)])
         self.assert_series_equal(result, expected)
 
-    @pytest.mark.skip(reason="Not Applicable", strict=True)
+    @pytest.mark.skip(reason="Not Applicable")
     def test_fillna_length_mismatch(self, data_missing):
         super().test_fillna_length_mismatch(data_missing)
 
diff --git a/pandas/tests/frame/test_combine_concat.py b/pandas/tests/frame/test_combine_concat.py
index 25c5222b5f03c..c60bb57625d75 100644
--- a/pandas/tests/frame/test_combine_concat.py
+++ b/pandas/tests/frame/test_combine_concat.py
@@ -77,8 +77,7 @@ def test_concat_multiple_tzs(self):
         [
             '2015-01-01',
             pytest.param(pd.NaT, marks=pytest.mark.xfail(
-                reason='GH23037 incorrect dtype when concatenating',
-                strict=True))])
+                reason='GH23037 incorrect dtype when concatenating'))])
     def test_concat_tz_NaT(self, t1):
         # GH 22796
         # Concating tz-aware multicolumn DataFrames
diff --git a/pandas/tests/frame/test_duplicates.py b/pandas/tests/frame/test_duplicates.py
index 3478d66b919a6..c9aff97bfa4b1 100644
--- a/pandas/tests/frame/test_duplicates.py
+++ b/pandas/tests/frame/test_duplicates.py
@@ -55,8 +55,7 @@ def test_duplicated_keep(keep, expected):
     tm.assert_series_equal(result, expected)
 
 
-@pytest.mark.xfail(reason="GH#21720; nan/None falsely considered equal",
-                   strict=True)
+@pytest.mark.xfail(reason="GH#21720; nan/None falsely considered equal")
 @pytest.mark.parametrize('keep, expected', [
     ('first', Series([False, False, True, False, True])),
     ('last', Series([True, True, False, False, False])),
diff --git a/pandas/tests/groupby/test_apply.py b/pandas/tests/groupby/test_apply.py
index 3bc5e51ca046a..8366f75a5795e 100644
--- a/pandas/tests/groupby/test_apply.py
+++ b/pandas/tests/groupby/test_apply.py
@@ -60,8 +60,7 @@ def test_apply_trivial():
 
 @pytest.mark.xfail(reason="GH#20066; function passed into apply "
                           "returns a DataFrame with the same index "
-                          "as the one to create GroupBy object.",
-                   strict=True)
+                          "as the one to create GroupBy object.")
 def test_apply_trivial_fail():
     # GH 20066
     # trivial apply fails if the constant dataframe has the same index
diff --git a/pandas/tests/groupby/test_categorical.py b/pandas/tests/groupby/test_categorical.py
index 14a09b83e5b7c..3692d34afcc03 100644
--- a/pandas/tests/groupby/test_categorical.py
+++ b/pandas/tests/groupby/test_categorical.py
@@ -206,7 +206,7 @@ def test_level_get_group(observed):
     assert_frame_equal(result, expected)
 
 
-@pytest.mark.xfail(PY37, reason="flaky on 3.7, xref gh-21636")
+@pytest.mark.xfail(PY37, reason="flaky on 3.7, xref gh-21636", strict=False)
 @pytest.mark.parametrize('ordered', [True, False])
 def test_apply(ordered):
     # GH 10138
diff --git a/pandas/tests/indexes/interval/test_astype.py b/pandas/tests/indexes/interval/test_astype.py
index 8bcd6ef5dcc5a..8406a8c458cca 100644
--- a/pandas/tests/indexes/interval/test_astype.py
+++ b/pandas/tests/indexes/interval/test_astype.py
@@ -92,7 +92,7 @@ def test_subtype_integer(self, subtype_start, subtype_end):
                                              closed=index.closed)
         tm.assert_index_equal(result, expected)
 
-    @pytest.mark.xfail(reason='GH#15832', strict=True)
+    @pytest.mark.xfail(reason='GH#15832')
     def test_subtype_integer_errors(self):
         # int64 -> uint64 fails with negative values
         index = interval_range(-10, 10)
@@ -130,7 +130,7 @@ def test_subtype_integer(self, subtype):
         with pytest.raises(ValueError, match=msg):
             index.insert(0, np.nan).astype(dtype)
 
-    @pytest.mark.xfail(reason='GH#15832', strict=True)
+    @pytest.mark.xfail(reason='GH#15832')
     def test_subtype_integer_errors(self):
         # float64 -> uint64 fails with negative values
         index = interval_range(-10.0, 10.0)
diff --git a/pandas/tests/indexes/multi/test_missing.py b/pandas/tests/indexes/multi/test_missing.py
index 73e6579cf7771..8ce33f100a6af 100644
--- a/pandas/tests/indexes/multi/test_missing.py
+++ b/pandas/tests/indexes/multi/test_missing.py
@@ -84,7 +84,7 @@ def test_nulls(idx):
         idx.isna()
 
 
-@pytest.mark.xfail(strict=True)
+@pytest.mark.xfail
 def test_hasnans_isnans(idx):
     # GH 11343, added tests for hasnans / isnans
     index = idx.copy()
diff --git a/pandas/tests/indexes/test_base.py b/pandas/tests/indexes/test_base.py
index 7028b5c23225c..fe7391ff15ebe 100644
--- a/pandas/tests/indexes/test_base.py
+++ b/pandas/tests/indexes/test_base.py
@@ -462,8 +462,7 @@ def test_constructor_overflow_int64(self):
             Index([np.iinfo(np.uint64).max - 1], dtype="int64")
 
     @pytest.mark.xfail(reason="see GH#21311: Index "
-                              "doesn't enforce dtype argument",
-                       strict=True)
+                              "doesn't enforce dtype argument")
     def test_constructor_cast(self):
         msg = "could not convert string to float"
         with pytest.raises(ValueError, match=msg):
@@ -1471,8 +1470,7 @@ def test_slice_float_locs(self):
         assert index2.slice_locs(8.5, 1.5) == (2, 6)
         assert index2.slice_locs(10.5, -1) == (0, n)
 
-    @pytest.mark.xfail(reason="Assertions were not correct - see GH#20915",
-                       strict=True)
+    @pytest.mark.xfail(reason="Assertions were not correct - see GH#20915")
     def test_slice_ints_with_floats_raises(self):
         # int slicing with floats
         # GH 4892, these are all TypeErrors
diff --git a/pandas/tests/indexing/test_coercion.py b/pandas/tests/indexing/test_coercion.py
index 2bc3aefcf7eb1..29b60d80750b2 100644
--- a/pandas/tests/indexing/test_coercion.py
+++ b/pandas/tests/indexing/test_coercion.py
@@ -903,11 +903,12 @@ def test_replace_series_datetime_tz(self):
 
     # TODO(jreback) commented out to only have a single xfail printed
     @pytest.mark.xfail(reason="different tz, "
-                       "currently mask_missing raises SystemError")
+                       "currently mask_missing raises SystemError",
+                       strict=False)
     # @pytest.mark.parametrize('how', ['dict', 'series'])
     # @pytest.mark.parametrize('to_key', [
-    #     'datetime64[ns]', 'datetime64[ns, UTC]',
-    #     'datetime64[ns, US/Eastern]'])
+    #    'datetime64[ns]', 'datetime64[ns, UTC]',
+    #    'datetime64[ns, US/Eastern]'])
     # @pytest.mark.parametrize('from_key', [
     #    'datetime64[ns]', 'datetime64[ns, UTC]',
     #    'datetime64[ns, US/Eastern]'])
diff --git a/pandas/tests/io/formats/test_to_csv.py b/pandas/tests/io/formats/test_to_csv.py
index 3792da4b29ef9..cba3f000b59c1 100644
--- a/pandas/tests/io/formats/test_to_csv.py
+++ b/pandas/tests/io/formats/test_to_csv.py
@@ -358,7 +358,7 @@ def test_to_csv_string_array_ascii(self):
             with open(path, 'r') as f:
                 assert f.read() == expected_ascii
 
-    @pytest.mark.xfail(strict=True)
+    @pytest.mark.xfail
     def test_to_csv_string_array_utf8(self):
         # GH 10813
         str_array = [{'names': ['foo', 'bar']}, {'names': ['baz', 'qux']}]
diff --git a/pandas/tests/io/formats/test_to_html.py b/pandas/tests/io/formats/test_to_html.py
index e7f87b236ef7f..ce9aca3a87c51 100644
--- a/pandas/tests/io/formats/test_to_html.py
+++ b/pandas/tests/io/formats/test_to_html.py
@@ -220,7 +220,7 @@ def test_to_html_truncate_multi_index(self, datapath):
         expected = expected_html(datapath, 'truncate_multi_index')
         assert result == expected
 
-    @pytest.mark.xfail(reason='GH22887 TypeError', strict=True)
+    @pytest.mark.xfail(reason='GH22887 TypeError')
     def test_to_html_truncate_multi_index_sparse_off(self, datapath):
         arrays = [['bar', 'bar', 'baz', 'baz', 'foo', 'foo', 'qux', 'qux'],
                   ['one', 'two', 'one', 'two', 'one', 'two', 'one', 'two']]
diff --git a/pandas/tests/io/json/test_json_table_schema.py b/pandas/tests/io/json/test_json_table_schema.py
index 0b4ff2c34297a..bb154de14611c 100644
--- a/pandas/tests/io/json/test_json_table_schema.py
+++ b/pandas/tests/io/json/test_json_table_schema.py
@@ -498,7 +498,7 @@ class TestTableOrientReader(object):
         None,
         "idx",
         pytest.param("index",
-                     marks=pytest.mark.xfail(strict=True)),
+                     marks=pytest.mark.xfail),
         'level_0'])
     @pytest.mark.parametrize("vals", [
         {'ints': [1, 2, 3, 4]},
@@ -508,7 +508,7 @@ class TestTableOrientReader(object):
         {'ordered_cats': pd.Series(pd.Categorical(['a', 'b', 'c', 'c'],
                                                   ordered=True))},
         pytest.param({'floats': [1., 2., 3., 4.]},
-                     marks=pytest.mark.xfail(strict=True)),
+                     marks=pytest.mark.xfail),
         {'floats': [1.1, 2.2, 3.3, 4.4]},
         {'bools': [True, False, False, True]}])
     def test_read_json_table_orient(self, index_nm, vals, recwarn):
@@ -566,7 +566,7 @@ def test_multiindex(self, index_names):
         tm.assert_frame_equal(df, result)
 
     @pytest.mark.parametrize("strict_check", [
-        pytest.param(True, marks=pytest.mark.xfail(strict=True)),
+        pytest.param(True, marks=pytest.mark.xfail),
         False
     ])
     def test_empty_frame_roundtrip(self, strict_check):
diff --git a/pandas/tests/io/parser/test_usecols.py b/pandas/tests/io/parser/test_usecols.py
index d2ec1cf49445f..068227908a285 100644
--- a/pandas/tests/io/parser/test_usecols.py
+++ b/pandas/tests/io/parser/test_usecols.py
@@ -520,7 +520,9 @@ def test_raises_on_usecols_names_mismatch(all_parsers, usecols,
         tm.assert_frame_equal(result, expected)
 
 
-@pytest.mark.xfail(reason="see gh-16469: buggy behavior")
+@pytest.mark.xfail(
+    reason="see gh-16469: works on the C engine but not the Python engine",
+    strict=False)
 @pytest.mark.parametrize("usecols", [["A", "C"], [0, 2]])
 def test_usecols_subset_names_mismatch_orig_columns(all_parsers, usecols):
     data = "a,b,c,d\n1,2,3,4\n5,6,7,8"
diff --git a/pandas/tests/io/test_excel.py b/pandas/tests/io/test_excel.py
index 7cc1f1899db98..564364ea01432 100644
--- a/pandas/tests/io/test_excel.py
+++ b/pandas/tests/io/test_excel.py
@@ -2389,8 +2389,7 @@ def check_called(func):
     pytest.param('xlwt',
                  marks=pytest.mark.xfail(reason='xlwt does not support '
                                                 'openpyxl-compatible '
-                                                'style dicts',
-                                         strict=True)),
+                                                'style dicts')),
     'xlsxwriter',
     'openpyxl',
 ])
diff --git a/pandas/tests/io/test_parquet.py b/pandas/tests/io/test_parquet.py
index 6024fccb15c76..5964c44a31f48 100644
--- a/pandas/tests/io/test_parquet.py
+++ b/pandas/tests/io/test_parquet.py
@@ -201,8 +201,7 @@ def test_options_get_engine(fp, pa):
 
 
 @pytest.mark.xfail(is_platform_windows() or is_platform_mac(),
-                   reason="reading pa metadata failing on Windows/mac",
-                   strict=True)
+                   reason="reading pa metadata failing on Windows/mac")
 def test_cross_engine_pa_fp(df_cross_compat, pa, fp):
     # cross-compat with differing reading/writing engines
 
@@ -404,7 +403,8 @@ def test_basic(self, pa, df_full):
         check_round_trip(df, pa)
 
     # TODO: This doesn't fail on all systems; track down which
-    @pytest.mark.xfail(reason="pyarrow fails on this (ARROW-1883)")
+    @pytest.mark.xfail(reason="pyarrow fails on this (ARROW-1883)",
+                       strict=False)
     def test_basic_subset_columns(self, pa, df_full):
         # GH18628
 
@@ -422,7 +422,6 @@ def test_duplicate_columns(self, pa):
                           columns=list('aaa')).copy()
         self.check_error_on_write(df, pa, ValueError)
 
-    @pytest.mark.xfail(reason="failing for pyarrow < 0.11.0")
     def test_unsupported(self, pa):
         # period
         df = pd.DataFrame({'a': pd.period_range('2013', freq='M', periods=3)})
diff --git a/pandas/tests/plotting/test_datetimelike.py b/pandas/tests/plotting/test_datetimelike.py
index b26fcf71306dd..1d6c8dc404d2b 100644
--- a/pandas/tests/plotting/test_datetimelike.py
+++ b/pandas/tests/plotting/test_datetimelike.py
@@ -1075,6 +1075,7 @@ def test_irreg_dtypes(self):
         _, ax = self.plt.subplots()
         _check_plot_works(df.plot, ax=ax)
 
+    @pytest.mark.xfail(reason="fails with py2.7.15", strict=False)
     @pytest.mark.slow
     def test_time(self):
         t = datetime(1, 1, 1, 3, 30, 0)
diff --git a/pandas/tests/plotting/test_frame.py b/pandas/tests/plotting/test_frame.py
index f5708b24d22b1..4e047cd44c1e2 100644
--- a/pandas/tests/plotting/test_frame.py
+++ b/pandas/tests/plotting/test_frame.py
@@ -488,8 +488,7 @@ def test_subplots_timeseries_y_axis(self):
             testdata.plot(y="text")
 
     @pytest.mark.xfail(reason='not support for period, categorical, '
-                              'datetime_mixed_tz',
-                       strict=True)
+                              'datetime_mixed_tz')
     def test_subplots_timeseries_y_axis_not_supported(self):
         """
         This test will fail for:
@@ -2557,6 +2556,7 @@ def test_errorbar_asymmetrical(self):
 
         tm.close()
 
+    # This XPASSES when tested with mpl == 3.0.1
     @td.xfail_if_mpl_2_2
     def test_table(self):
         df = DataFrame(np.random.rand(10, 3),
diff --git a/pandas/tests/plotting/test_misc.py b/pandas/tests/plotting/test_misc.py
index 1f0a0d6bfee95..9ae3e7fc423f4 100644
--- a/pandas/tests/plotting/test_misc.py
+++ b/pandas/tests/plotting/test_misc.py
@@ -61,6 +61,7 @@ def test_bootstrap_plot(self):
 @td.skip_if_no_mpl
 class TestDataFramePlots(TestPlotBase):
 
+    # This XPASSES when tested with mpl == 3.0.1
     @td.xfail_if_mpl_2_2
     @td.skip_if_no_scipy
     def test_scatter_matrix_axis(self):
diff --git a/pandas/tests/plotting/test_series.py b/pandas/tests/plotting/test_series.py
index e6519c7db7a7b..cc8aa2018b1a0 100644
--- a/pandas/tests/plotting/test_series.py
+++ b/pandas/tests/plotting/test_series.py
@@ -771,6 +771,7 @@ def test_errorbar_plot(self):
         with pytest.raises((TypeError, ValueError)):
             s.plot(yerr=s_err)
 
+    # This XPASSES when tested with mpl == 3.0.1
     @td.xfail_if_mpl_2_2
     def test_table(self):
         _check_plot_works(self.series.plot, table=True)
diff --git a/pandas/tests/reshape/test_pivot.py b/pandas/tests/reshape/test_pivot.py
index 69572f75fea1b..b3dd94b49e3a3 100644
--- a/pandas/tests/reshape/test_pivot.py
+++ b/pandas/tests/reshape/test_pivot.py
@@ -489,8 +489,7 @@ def test_pivot_with_list_like_values_nans(self, values, method):
         tm.assert_frame_equal(result, expected)
 
     @pytest.mark.xfail(reason='MultiIndexed unstack with tuple names fails'
-                              'with KeyError GH#19966',
-                       strict=True)
+                              'with KeyError GH#19966')
     @pytest.mark.parametrize('method', [True, False])
     def test_pivot_with_multiindex(self, method):
         # issue #17160
@@ -616,8 +615,7 @@ def test_margins_dtype(self):
         tm.assert_frame_equal(expected, result)
 
     @pytest.mark.xfail(reason='GH#17035 (len of floats is casted back to '
-                              'floats)',
-                       strict=True)
+                              'floats)')
     def test_margins_dtype_len(self):
         mi_val = list(product(['bar', 'foo'], ['one', 'two'])) + [('All', '')]
         mi = MultiIndex.from_tuples(mi_val, names=('A', 'B'))
@@ -1102,8 +1100,7 @@ def test_pivot_table_margins_name_with_aggfunc_list(self):
         tm.assert_frame_equal(table, expected)
 
     @pytest.mark.xfail(reason='GH#17035 (np.mean of ints is casted back to '
-                              'ints)',
-                       strict=True)
+                              'ints)')
     def test_categorical_margins(self, observed):
         # GH 10989
         df = pd.DataFrame({'x': np.arange(8),
@@ -1118,8 +1115,7 @@ def test_categorical_margins(self, observed):
         tm.assert_frame_equal(table, expected)
 
     @pytest.mark.xfail(reason='GH#17035 (np.mean of ints is casted back to '
-                              'ints)',
-                       strict=True)
+                              'ints)')
     def test_categorical_margins_category(self, observed):
         df = pd.DataFrame({'x': np.arange(8),
                            'y': np.arange(8) // 4,
diff --git a/pandas/tests/scalar/period/test_asfreq.py b/pandas/tests/scalar/period/test_asfreq.py
index 064d1a96878c2..24f2ed88936b7 100644
--- a/pandas/tests/scalar/period/test_asfreq.py
+++ b/pandas/tests/scalar/period/test_asfreq.py
@@ -34,8 +34,7 @@ def test_asfreq_near_zero_weekly(self):
         assert week2.asfreq('D', 'S') <= per2
 
     @pytest.mark.xfail(reason='GH#19643 period_helper asfreq functions fail '
-                              'to check for overflows',
-                       strict=True)
+                              'to check for overflows')
     def test_to_timestamp_out_of_bounds(self):
         # GH#19643, currently gives Timestamp('1754-08-30 22:43:41.128654848')
         per = Period('0001-01-01', freq='B')
diff --git a/pandas/tests/scalar/period/test_period.py b/pandas/tests/scalar/period/test_period.py
index 4d3aa1109c120..ddb4f89738f98 100644
--- a/pandas/tests/scalar/period/test_period.py
+++ b/pandas/tests/scalar/period/test_period.py
@@ -1468,7 +1468,8 @@ def test_period_immutable():
 
 
 # TODO: This doesn't fail on all systems; track down which
-@pytest.mark.xfail(reason="Parses as Jan 1, 0007 on some systems")
+@pytest.mark.xfail(reason="Parses as Jan 1, 0007 on some systems",
+                   strict=False)
 def test_small_year_parsing():
     per1 = Period('0001-01-07', 'D')
     assert per1.year == 1
diff --git a/pandas/tests/series/test_analytics.py b/pandas/tests/series/test_analytics.py
index 89d76abaf5f82..6e40324c67b59 100644
--- a/pandas/tests/series/test_analytics.py
+++ b/pandas/tests/series/test_analytics.py
@@ -2098,7 +2098,7 @@ def test_value_counts_with_nan(self):
         "dtype",
         ["int_", "uint", "float_", "unicode_", "timedelta64[h]",
          pytest.param("datetime64[D]",
-                      marks=pytest.mark.xfail(reason="GH#7996", strict=True))]
+                      marks=pytest.mark.xfail(reason="GH#7996"))]
     )
     @pytest.mark.parametrize("is_ordered", [True, False])
     def test_drop_duplicates_categorical_non_bool(self, dtype, is_ordered):
diff --git a/pandas/tests/series/test_period.py b/pandas/tests/series/test_period.py
index ce620db8d9c1b..0a86bb0b67797 100644
--- a/pandas/tests/series/test_period.py
+++ b/pandas/tests/series/test_period.py
@@ -64,8 +64,7 @@ def test_between(self):
     # ---------------------------------------------------------------------
     # NaT support
 
-    @pytest.mark.xfail(reason="PeriodDtype Series not supported yet",
-                       strict=True)
+    @pytest.mark.xfail(reason="PeriodDtype Series not supported yet")
     def test_NaT_scalar(self):
         series = Series([0, 1000, 2000, pd._libs.iNaT], dtype='period[D]')
 
@@ -75,8 +74,7 @@ def test_NaT_scalar(self):
         series[2] = val
         assert pd.isna(series[2])
 
-    @pytest.mark.xfail(reason="PeriodDtype Series not supported yet",
-                       strict=True)
+    @pytest.mark.xfail(reason="PeriodDtype Series not supported yet")
     def test_NaT_cast(self):
         result = Series([np.nan]).astype('period[D]')
         expected = Series([pd.NaT])
diff --git a/pandas/tests/series/test_rank.py b/pandas/tests/series/test_rank.py
index 72d05cb4839ef..da414a577ae0b 100644
--- a/pandas/tests/series/test_rank.py
+++ b/pandas/tests/series/test_rank.py
@@ -222,8 +222,7 @@ def test_rank_signature(self):
                      'int64',
                      marks=pytest.mark.xfail(
                          reason="iNaT is equivalent to minimum value of dtype"
-                                "int64 pending issue GH#16674",
-                         strict=True)),
+                                "int64 pending issue GH#16674")),
         ([NegInfinity(), '1', 'A', 'BA', 'Ba', 'C', Infinity()],
          'object')
     ])
diff --git a/pandas/tests/sparse/frame/test_analytics.py b/pandas/tests/sparse/frame/test_analytics.py
index 54e3ddbf2f1cf..2d9ccaa059a8c 100644
--- a/pandas/tests/sparse/frame/test_analytics.py
+++ b/pandas/tests/sparse/frame/test_analytics.py
@@ -4,8 +4,7 @@
 from pandas.util import testing as tm
 
 
-@pytest.mark.xfail(reason='Wrong SparseBlock initialization (GH#17386)',
-                   strict=True)
+@pytest.mark.xfail(reason='Wrong SparseBlock initialization (GH#17386)')
 def test_quantile():
     # GH 17386
     data = [[1, 1], [2, 10], [3, 100], [np.nan, np.nan]]
@@ -22,8 +21,7 @@ def test_quantile():
     tm.assert_sp_series_equal(result, sparse_expected)
 
 
-@pytest.mark.xfail(reason='Wrong SparseBlock initialization (GH#17386)',
-                   strict=True)
+@pytest.mark.xfail(reason='Wrong SparseBlock initialization (GH#17386)')
 def test_quantile_multi():
     # GH 17386
     data = [[1, 1], [2, 10], [3, 100], [np.nan, np.nan]]
diff --git a/pandas/tests/sparse/frame/test_frame.py b/pandas/tests/sparse/frame/test_frame.py
index f799eab2f6406..21100e3c3ffeb 100644
--- a/pandas/tests/sparse/frame/test_frame.py
+++ b/pandas/tests/sparse/frame/test_frame.py
@@ -1140,7 +1140,7 @@ def test_combine_first(self, float_frame):
 
         tm.assert_sp_frame_equal(result, expected)
 
-    @pytest.mark.xfail(reason="No longer supported.", strict=True)
+    @pytest.mark.xfail(reason="No longer supported.")
     def test_combine_first_with_dense(self):
         # We could support this if we allow
         # pd.core.dtypes.cast.find_common_type to special case SparseDtype
@@ -1196,8 +1196,7 @@ def test_as_blocks(self):
         tm.assert_frame_equal(df_blocks['Sparse[float64, nan]'], df)
 
     @pytest.mark.xfail(reason='nan column names in _init_dict problematic '
-                              '(GH#16894)',
-                       strict=True)
+                              '(GH#16894)')
     def test_nan_columnname(self):
         # GH 8822
         nan_colname = DataFrame(Series(1.0, index=[0]), columns=[nan])
@@ -1314,8 +1313,7 @@ def test_numpy_func_call(self, float_frame):
         for func in funcs:
             getattr(np, func)(float_frame)
 
-    @pytest.mark.xfail(reason='Wrong SparseBlock initialization (GH 17386)',
-                       strict=True)
+    @pytest.mark.xfail(reason='Wrong SparseBlock initialization (GH 17386)')
     def test_quantile(self):
         # GH 17386
         data = [[1, 1], [2, 10], [3, 100], [nan, nan]]
@@ -1331,8 +1329,7 @@ def test_quantile(self):
         tm.assert_series_equal(result, dense_expected)
         tm.assert_sp_series_equal(result, sparse_expected)
 
-    @pytest.mark.xfail(reason='Wrong SparseBlock initialization (GH 17386)',
-                       strict=True)
+    @pytest.mark.xfail(reason='Wrong SparseBlock initialization (GH 17386)')
     def test_quantile_multi(self):
         # GH 17386
         data = [[1, 1], [2, 10], [3, 100], [nan, nan]]
diff --git a/pandas/tests/sparse/frame/test_indexing.py b/pandas/tests/sparse/frame/test_indexing.py
index 607eb2da6ded0..e4ca3b90ff8d0 100644
--- a/pandas/tests/sparse/frame/test_indexing.py
+++ b/pandas/tests/sparse/frame/test_indexing.py
@@ -18,8 +18,7 @@
         [np.nan, np.nan]
     ]
 ])
-@pytest.mark.xfail(reason='Wrong SparseBlock initialization (GH#17386)',
-                   strict=True)
+@pytest.mark.xfail(reason='Wrong SparseBlock initialization (GH#17386)')
 def test_where_with_numeric_data(data):
     # GH 17386
     lower_bound = 1.5
@@ -52,8 +51,7 @@ def test_where_with_numeric_data(data):
     0.1,
     100.0 + 100.0j
 ])
-@pytest.mark.xfail(reason='Wrong SparseBlock initialization (GH#17386)',
-                   strict=True)
+@pytest.mark.xfail(reason='Wrong SparseBlock initialization (GH#17386)')
 def test_where_with_numeric_data_and_other(data, other):
     # GH 17386
     lower_bound = 1.5
@@ -70,8 +68,7 @@ def test_where_with_numeric_data_and_other(data, other):
     tm.assert_sp_frame_equal(result, sparse_expected)
 
 
-@pytest.mark.xfail(reason='Wrong SparseBlock initialization (GH#17386)',
-                   strict=True)
+@pytest.mark.xfail(reason='Wrong SparseBlock initialization (GH#17386)')
 def test_where_with_bool_data():
     # GH 17386
     data = [[False, False], [True, True], [False, False]]
@@ -94,8 +91,7 @@ def test_where_with_bool_data():
     0.1,
     100.0 + 100.0j
 ])
-@pytest.mark.xfail(reason='Wrong SparseBlock initialization (GH#17386)',
-                   strict=True)
+@pytest.mark.xfail(reason='Wrong SparseBlock initialization (GH#17386)')
 def test_where_with_bool_data_and_other(other):
     # GH 17386
     data = [[False, False], [True, True], [False, False]]
diff --git a/pandas/tests/sparse/series/test_indexing.py b/pandas/tests/sparse/series/test_indexing.py
index 998285d933492..989cf3b974560 100644
--- a/pandas/tests/sparse/series/test_indexing.py
+++ b/pandas/tests/sparse/series/test_indexing.py
@@ -18,8 +18,7 @@
         np.nan, np.nan
     ]
 ])
-@pytest.mark.xfail(reason='Wrong SparseBlock initialization (GH#17386)',
-                   strict=True)
+@pytest.mark.xfail(reason='Wrong SparseBlock initialization (GH#17386)')
 def test_where_with_numeric_data(data):
     # GH 17386
     lower_bound = 1.5
@@ -70,8 +69,7 @@ def test_where_with_numeric_data_and_other(data, other):
     tm.assert_sp_series_equal(result, sparse_expected)
 
 
-@pytest.mark.xfail(reason='Wrong SparseBlock initialization (GH#17386)',
-                   strict=True)
+@pytest.mark.xfail(reason='Wrong SparseBlock initialization (GH#17386)')
 def test_where_with_bool_data():
     # GH 17386
     data = [False, False, True, True, False, False]
diff --git a/pandas/tests/test_downstream.py b/pandas/tests/test_downstream.py
index abcfa4b320b22..1d17b514a5b67 100644
--- a/pandas/tests/test_downstream.py
+++ b/pandas/tests/test_downstream.py
@@ -101,7 +101,7 @@ def test_pandas_gbq(df):
     pandas_gbq = import_module('pandas_gbq')  # noqa
 
 
-@pytest.mark.xfail(reason="0.7.0 pending", strict=True)
+@pytest.mark.xfail(reason="0.7.0 pending")
 @tm.network
 def test_pandas_datareader():
 
diff --git a/pandas/tests/test_window.py b/pandas/tests/test_window.py
index 31ea5c11f5bd1..b53aca2c9852b 100644
--- a/pandas/tests/test_window.py
+++ b/pandas/tests/test_window.py
@@ -695,8 +695,7 @@ def test_numpy_compat(self, method):
         'expander',
         [1, pytest.param('ls', marks=pytest.mark.xfail(
                          reason='GH#16425 expanding with '
-                                'offset not supported',
-                         strict=True))])
+                                'offset not supported'))])
     def test_empty_df_expanding(self, expander):
         # GH 15819 Verifies that datetime and integer expanding windows can be
         # applied to empty DataFrames
diff --git a/pandas/tests/tseries/offsets/test_offsets_properties.py b/pandas/tests/tseries/offsets/test_offsets_properties.py
index a1b68d34f4e18..cd5f2a2a25e58 100644
--- a/pandas/tests/tseries/offsets/test_offsets_properties.py
+++ b/pandas/tests/tseries/offsets/test_offsets_properties.py
@@ -72,7 +72,7 @@ def test_on_offset_implementations(dt, offset):
     assert offset.onOffset(dt) == (compare == dt)
 
 
-@pytest.mark.xfail(strict=True)
+@pytest.mark.xfail
 @given(gen_yqm_offset, gen_date_range)
 def test_apply_index_implementations(offset, rng):
     # offset.apply_index(dti)[i] should match dti[i] + offset
@@ -94,7 +94,7 @@ def test_apply_index_implementations(offset, rng):
     # TODO: Check randomly assorted entries, not just first/last
 
 
-@pytest.mark.xfail(strict=True)
+@pytest.mark.xfail
 @given(gen_yqm_offset)
 def test_shift_across_dst(offset):
     # GH#18319 check that 1) timezone is correctly normalized and
diff --git a/pandas/util/_test_decorators.py b/pandas/util/_test_decorators.py
index 3f8332ade4487..0331661c3131f 100644
--- a/pandas/util/_test_decorators.py
+++ b/pandas/util/_test_decorators.py
@@ -158,7 +158,8 @@ def decorated_func(func):
 skip_if_mpl = pytest.mark.skipif(not _skip_if_no_mpl(),
                                  reason="matplotlib is present")
 xfail_if_mpl_2_2 = pytest.mark.xfail(_skip_if_mpl_2_2(),
-                                     reason="matplotlib 2.2")
+                                     reason="matplotlib 2.2",
+                                     strict=False)
 skip_if_32bit = pytest.mark.skipif(is_platform_32bit(),
                                    reason="skipping for 32 bit")
 skip_if_windows = pytest.mark.skipif(is_platform_windows(),
diff --git a/setup.cfg b/setup.cfg
index b9994e9ea0b2c..6c1f09555df1c 100644
--- a/setup.cfg
+++ b/setup.cfg
@@ -106,6 +106,7 @@ markers =
     clipboard: mark a pd.read_clipboard test
 doctest_optionflags = NORMALIZE_WHITESPACE IGNORE_EXCEPTION_DETAIL
 addopts = --strict-data-files
+xfail_strict = True
 
 [coverage:run]
 branch = False