diff --git a/pandas/tests/arithmetic/test_datetime64.py b/pandas/tests/arithmetic/test_datetime64.py
index a62bebec61ada..fae0a7a0d95f8 100644
--- a/pandas/tests/arithmetic/test_datetime64.py
+++ b/pandas/tests/arithmetic/test_datetime64.py
@@ -414,7 +414,7 @@ def test_dti_cmp_datetimelike(self, other, tz_naive_fixture):
         dti = date_range("2016-01-01", periods=2, tz=tz)
         if tz is not None:
             if isinstance(other, np.datetime64):
-                pytest.skip("no tzaware version available")
+                pytest.skip(f"{type(other).__name__} is not tz aware")
             other = localize_pydatetime(other, dti.tzinfo)
 
         result = dti == other
diff --git a/pandas/tests/arithmetic/test_numeric.py b/pandas/tests/arithmetic/test_numeric.py
index 8d49171cc43f3..c2fba3c775de9 100644
--- a/pandas/tests/arithmetic/test_numeric.py
+++ b/pandas/tests/arithmetic/test_numeric.py
@@ -384,7 +384,7 @@ def test_divmod_zero(self, zero, numeric_idx):
     def test_div_negative_zero(self, zero, numeric_idx, op):
         # Check that -1 / -0.0 returns np.inf, not -np.inf
         if numeric_idx.dtype == np.uint64:
-            pytest.skip(f"Not relevant for {numeric_idx.dtype}")
+            pytest.skip(f"Div by negative 0 not relevant for {numeric_idx.dtype}")
         idx = numeric_idx - 3
 
         expected = Index([-np.inf, -np.inf, -np.inf, np.nan, np.inf], dtype=np.float64)
diff --git a/pandas/tests/computation/test_eval.py b/pandas/tests/computation/test_eval.py
index 40c3069dfeebf..fe49446424de1 100644
--- a/pandas/tests/computation/test_eval.py
+++ b/pandas/tests/computation/test_eval.py
@@ -63,7 +63,7 @@
                     reason=f"numexpr enabled->{USE_NUMEXPR}, "
                     f"installed->{NUMEXPR_INSTALLED}",
                 ),
-                td.skip_if_no_ne,
+                td.skip_if_no("numexpr"),
             ],
         )
         for engine in ENGINES
@@ -1687,14 +1687,14 @@ def test_empty_globals(self, engine, parser):
             pd.eval(e, engine=engine, parser=parser, global_dict={})
 
 
-@td.skip_if_no_ne
+@td.skip_if_no("numexpr")
 def test_invalid_engine():
     msg = "Invalid engine 'asdf' passed"
     with pytest.raises(KeyError, match=msg):
         pd.eval("x + y", local_dict={"x": 1, "y": 2}, engine="asdf")
 
 
-@td.skip_if_no_ne
+@td.skip_if_no("numexpr")
 @pytest.mark.parametrize(
     ("use_numexpr", "expected"),
     (
@@ -1711,7 +1711,7 @@ def test_numexpr_option_respected(use_numexpr, expected):
         assert result == expected
 
 
-@td.skip_if_no_ne
+@td.skip_if_no("numexpr")
 def test_numexpr_option_incompatible_op():
     # GH 32556
     with pd.option_context("compute.use_numexpr", False):
@@ -1723,7 +1723,7 @@ def test_numexpr_option_incompatible_op():
         tm.assert_frame_equal(result, expected)
 
 
-@td.skip_if_no_ne
+@td.skip_if_no("numexpr")
 def test_invalid_parser():
     msg = "Invalid parser 'asdf' passed"
     with pytest.raises(KeyError, match=msg):
diff --git a/pandas/tests/dtypes/test_common.py b/pandas/tests/dtypes/test_common.py
index 8e99c074a2c55..d946b8da01fad 100644
--- a/pandas/tests/dtypes/test_common.py
+++ b/pandas/tests/dtypes/test_common.py
@@ -196,7 +196,7 @@ def test_is_object():
 
 
 @pytest.mark.parametrize(
-    "check_scipy", [False, pytest.param(True, marks=td.skip_if_no_scipy)]
+    "check_scipy", [False, pytest.param(True, marks=td.skip_if_no("scipy"))]
 )
 def test_is_sparse(check_scipy):
     msg = "is_sparse is deprecated"
@@ -632,7 +632,7 @@ def test_is_bool_dtype_numpy_error():
 
 
 @pytest.mark.parametrize(
-    "check_scipy", [False, pytest.param(True, marks=td.skip_if_no_scipy)]
+    "check_scipy", [False, pytest.param(True, marks=td.skip_if_no("scipy"))]
 )
 def test_is_extension_array_dtype(check_scipy):
     assert not com.is_extension_array_dtype([1, 2, 3])
diff --git a/pandas/tests/extension/base/dim2.py b/pandas/tests/extension/base/dim2.py
index 12006248b1db3..132cda5a94ed0 100644
--- a/pandas/tests/extension/base/dim2.py
+++ b/pandas/tests/extension/base/dim2.py
@@ -29,7 +29,7 @@ def skip_if_doesnt_support_2d(self, dtype, request):
             test_func = node._obj
             if test_func.__qualname__.startswith("Dim2CompatTests"):
                 # TODO: is there a less hacky way of checking this?
-                pytest.skip("Test is only for EAs that support 2D.")
+                pytest.skip(f"{dtype} does not support 2D.")
 
     def test_transpose(self, data):
         arr2d = data.repeat(2).reshape(-1, 2)
diff --git a/pandas/tests/extension/base/interface.py b/pandas/tests/extension/base/interface.py
index f19561e5a862b..08e385c9389d9 100644
--- a/pandas/tests/extension/base/interface.py
+++ b/pandas/tests/extension/base/interface.py
@@ -103,7 +103,7 @@ def test_copy(self, data):
         result = data.copy()
 
         if data.dtype._is_immutable:
-            pytest.skip("test_copy assumes mutability")
+            pytest.skip(f"test_copy assumes mutability and {data.dtype} is immutable")
 
         data[1] = data[0]
         assert result[1] != result[0]
@@ -118,7 +118,7 @@ def test_view(self, data):
         assert type(result) == type(data)
 
         if data.dtype._is_immutable:
-            pytest.skip("test_view assumes mutability")
+            pytest.skip(f"test_view assumes mutability and {data.dtype} is immutable")
 
         result[1] = result[0]
         assert data[1] == data[0]
diff --git a/pandas/tests/extension/base/reduce.py b/pandas/tests/extension/base/reduce.py
index 4477ba17c1683..6ea1b3a6fbe9d 100644
--- a/pandas/tests/extension/base/reduce.py
+++ b/pandas/tests/extension/base/reduce.py
@@ -120,7 +120,7 @@ def test_reduce_frame(self, data, all_numeric_reductions, skipna):
         op_name = all_numeric_reductions
         ser = pd.Series(data)
         if not is_numeric_dtype(ser.dtype):
-            pytest.skip("not numeric dtype")
+            pytest.skip(f"{ser.dtype} is not numeric dtype")
 
         if op_name in ["count", "kurt", "sem"]:
             pytest.skip(f"{op_name} not an array method")
diff --git a/pandas/tests/extension/base/reshaping.py b/pandas/tests/extension/base/reshaping.py
index c320ccfe4c548..227a1b76088cb 100644
--- a/pandas/tests/extension/base/reshaping.py
+++ b/pandas/tests/extension/base/reshaping.py
@@ -343,7 +343,7 @@ def test_ravel(self, data):
         assert type(result) == type(data)
 
         if data.dtype._is_immutable:
-            pytest.skip("test_ravel assumes mutability")
+            pytest.skip(f"test_ravel assumes mutability and {data.dtype} is immutable")
 
         # Check that we have a view, not a copy
         result[0] = result[1]
@@ -360,7 +360,9 @@ def test_transpose(self, data):
         assert result.shape == data.shape[::-1]
 
         if data.dtype._is_immutable:
-            pytest.skip("test_transpose assumes mutability")
+            pytest.skip(
+                f"test_transpose assumes mutability and {data.dtype} is immutable"
+            )
 
         # Check that we have a view, not a copy
         result[0] = result[1]
diff --git a/pandas/tests/extension/base/setitem.py b/pandas/tests/extension/base/setitem.py
index b8ede450b3c1a..2e2cbf5bf0d83 100644
--- a/pandas/tests/extension/base/setitem.py
+++ b/pandas/tests/extension/base/setitem.py
@@ -43,7 +43,7 @@ def skip_if_immutable(self, dtype, request):
                 # This fixture is auto-used, but we want to not-skip
                 # test_is_immutable.
                 return
-            pytest.skip("__setitem__ test not applicable with immutable dtype")
+            pytest.skip(f"__setitem__ test not applicable with immutable dtype {dtype}")
 
     def test_is_immutable(self, data):
         if data.dtype._is_immutable:
diff --git a/pandas/tests/extension/conftest.py b/pandas/tests/extension/conftest.py
index 30b10541c8f66..27f6eabfd126e 100644
--- a/pandas/tests/extension/conftest.py
+++ b/pandas/tests/extension/conftest.py
@@ -37,7 +37,7 @@ def data_for_twos(dtype):
     if not (dtype._is_numeric or dtype.kind == "m"):
         # Object-dtypes may want to allow this, but for the most part
         #  only numeric and timedelta-like dtypes will need to implement this.
-        pytest.skip("Not a numeric dtype")
+        pytest.skip(f"{dtype} is not a numeric dtype")
 
     raise NotImplementedError
 
diff --git a/pandas/tests/extension/test_interval.py b/pandas/tests/extension/test_interval.py
index f37ac4b289852..b427c12d2d40c 100644
--- a/pandas/tests/extension/test_interval.py
+++ b/pandas/tests/extension/test_interval.py
@@ -56,7 +56,7 @@ def data_missing():
 
 @pytest.fixture
 def data_for_twos():
-    pytest.skip("Not a numeric dtype")
+    pytest.skip("Interval is not a numeric dtype")
 
 
 @pytest.fixture
diff --git a/pandas/tests/extension/test_numpy.py b/pandas/tests/extension/test_numpy.py
index 04b25fd0566da..f1939ea174841 100644
--- a/pandas/tests/extension/test_numpy.py
+++ b/pandas/tests/extension/test_numpy.py
@@ -151,7 +151,7 @@ def data_for_grouping(allow_in_pandas, dtype):
 @pytest.fixture
 def data_for_twos(dtype):
     if dtype.kind == "O":
-        pytest.skip("Not a numeric dtype")
+        pytest.skip(f"{dtype} is not a numeric dtype")
     arr = np.ones(100) * 2
     return NumpyExtensionArray._from_sequence(arr, dtype=dtype)
 
@@ -323,7 +323,7 @@ def check_reduce(self, ser: pd.Series, op_name: str, skipna: bool):
             expected = exp_op(skipna=skipna)
         tm.assert_almost_equal(result, expected)
 
-    @pytest.mark.skip("tests not written yet")
+    @pytest.mark.skip("TODO: tests not written yet")
     @pytest.mark.parametrize("skipna", [True, False])
     def test_reduce_frame(self, data, all_numeric_reductions, skipna):
         pass
diff --git a/pandas/tests/extension/test_string.py b/pandas/tests/extension/test_string.py
index 840dd1057745f..8c5098a53adfc 100644
--- a/pandas/tests/extension/test_string.py
+++ b/pandas/tests/extension/test_string.py
@@ -26,22 +26,21 @@
 from pandas.tests.extension import base
 
 
-def split_array(arr):
-    if arr.dtype.storage != "pyarrow":
-        pytest.skip("only applicable for pyarrow chunked array n/a")
-
-    def _split_array(arr):
-        import pyarrow as pa
-
-        arrow_array = arr._pa_array
-        split = len(arrow_array) // 2
-        arrow_array = pa.chunked_array(
-            [*arrow_array[:split].chunks, *arrow_array[split:].chunks]
-        )
-        assert arrow_array.num_chunks == 2
-        return type(arr)(arrow_array)
-
-    return _split_array(arr)
+def maybe_split_array(arr, chunked):
+    if not chunked:
+        return arr
+    elif arr.dtype.storage != "pyarrow":
+        return arr
+
+    pa = pytest.importorskip("pyarrow")
+
+    arrow_array = arr._pa_array
+    split = len(arrow_array) // 2
+    arrow_array = pa.chunked_array(
+        [*arrow_array[:split].chunks, *arrow_array[split:].chunks]
+    )
+    assert arrow_array.num_chunks == 2
+    return type(arr)(arrow_array)
 
 
 @pytest.fixture(params=[True, False])
@@ -61,26 +60,26 @@ def data(dtype, chunked):
         strings = np.random.default_rng(2).choice(list(string.ascii_letters), size=100)
 
     arr = dtype.construct_array_type()._from_sequence(strings)
-    return split_array(arr) if chunked else arr
+    return maybe_split_array(arr, chunked)
 
 
 @pytest.fixture
 def data_missing(dtype, chunked):
     """Length 2 array with [NA, Valid]"""
     arr = dtype.construct_array_type()._from_sequence([pd.NA, "A"])
-    return split_array(arr) if chunked else arr
+    return maybe_split_array(arr, chunked)
 
 
 @pytest.fixture
 def data_for_sorting(dtype, chunked):
     arr = dtype.construct_array_type()._from_sequence(["B", "C", "A"])
-    return split_array(arr) if chunked else arr
+    return maybe_split_array(arr, chunked)
 
 
 @pytest.fixture
 def data_missing_for_sorting(dtype, chunked):
     arr = dtype.construct_array_type()._from_sequence(["B", pd.NA, "A"])
-    return split_array(arr) if chunked else arr
+    return maybe_split_array(arr, chunked)
 
 
 @pytest.fixture
@@ -88,7 +87,7 @@ def data_for_grouping(dtype, chunked):
     arr = dtype.construct_array_type()._from_sequence(
         ["B", "B", pd.NA, pd.NA, "A", "A", "B", "C"]
     )
-    return split_array(arr) if chunked else arr
+    return maybe_split_array(arr, chunked)
 
 
 class TestDtype(base.BaseDtypeTests):
diff --git a/pandas/tests/frame/methods/test_interpolate.py b/pandas/tests/frame/methods/test_interpolate.py
index bb12d7e202e09..1ddab99947163 100644
--- a/pandas/tests/frame/methods/test_interpolate.py
+++ b/pandas/tests/frame/methods/test_interpolate.py
@@ -322,7 +322,7 @@ def test_rowwise_alt(self):
         # TODO: assert something?
 
     @pytest.mark.parametrize(
-        "check_scipy", [False, pytest.param(True, marks=td.skip_if_no_scipy)]
+        "check_scipy", [False, pytest.param(True, marks=td.skip_if_no("scipy"))]
     )
     def test_interp_leading_nans(self, check_scipy):
         df = DataFrame(
diff --git a/pandas/tests/frame/test_query_eval.py b/pandas/tests/frame/test_query_eval.py
index 72e8236159bda..2e8c5258547c3 100644
--- a/pandas/tests/frame/test_query_eval.py
+++ b/pandas/tests/frame/test_query_eval.py
@@ -27,7 +27,8 @@ def parser(request):
 
 
 @pytest.fixture(
-    params=["python", pytest.param("numexpr", marks=td.skip_if_no_ne)], ids=lambda x: x
+    params=["python", pytest.param("numexpr", marks=td.skip_if_no("numexpr"))],
+    ids=lambda x: x,
 )
 def engine(request):
     return request.param
@@ -35,7 +36,7 @@ def engine(request):
 
 def skip_if_no_pandas_parser(parser):
     if parser != "pandas":
-        pytest.skip(f"cannot evaluate with parser {repr(parser)}")
+        pytest.skip(f"cannot evaluate with parser={parser}")
 
 
 class TestCompat:
@@ -387,7 +388,7 @@ def to_series(mi, level):
                 raise AssertionError("object must be a Series or Index")
 
 
-@td.skip_if_no_ne
+@td.skip_if_no("numexpr")
 class TestDataFrameQueryNumExprPandas:
     @pytest.fixture
     def engine(self):
@@ -765,7 +766,7 @@ def test_method_calls_in_query(self, engine, parser):
         tm.assert_frame_equal(result, expected)
 
 
-@td.skip_if_no_ne
+@td.skip_if_no("numexpr")
 class TestDataFrameQueryNumExprPython(TestDataFrameQueryNumExprPandas):
     @pytest.fixture
     def engine(self):
diff --git a/pandas/tests/frame/test_reductions.py b/pandas/tests/frame/test_reductions.py
index 25d6dcaf53bfb..fc7c1b0f01fed 100644
--- a/pandas/tests/frame/test_reductions.py
+++ b/pandas/tests/frame/test_reductions.py
@@ -165,8 +165,8 @@ class TestDataFrameAnalytics:
             "var",
             "std",
             "sem",
-            pytest.param("skew", marks=td.skip_if_no_scipy),
-            pytest.param("kurt", marks=td.skip_if_no_scipy),
+            pytest.param("skew", marks=td.skip_if_no("scipy")),
+            pytest.param("kurt", marks=td.skip_if_no("scipy")),
         ],
     )
     def test_stat_op_api_float_string_frame(self, float_string_frame, axis, opname):
@@ -219,8 +219,8 @@ def test_stat_op_api_float_string_frame(self, float_string_frame, axis, opname):
             "var",
             "std",
             "sem",
-            pytest.param("skew", marks=td.skip_if_no_scipy),
-            pytest.param("kurt", marks=td.skip_if_no_scipy),
+            pytest.param("skew", marks=td.skip_if_no("scipy")),
+            pytest.param("kurt", marks=td.skip_if_no("scipy")),
         ],
     )
     def test_stat_op_api_float_frame(self, float_frame, axis, opname):
diff --git a/pandas/tests/groupby/test_reductions.py b/pandas/tests/groupby/test_reductions.py
index e5836ce1e61c9..30ca3110ac2fa 100644
--- a/pandas/tests/groupby/test_reductions.py
+++ b/pandas/tests/groupby/test_reductions.py
@@ -933,7 +933,7 @@ def scipy_sem(*args, **kwargs):
         ("first", lambda x: x.iloc[0]),
         ("last", lambda x: x.iloc[-1]),
         ("count", np.size),
-        pytest.param("sem", scipy_sem, marks=td.skip_if_no_scipy),
+        pytest.param("sem", scipy_sem, marks=td.skip_if_no("scipy")),
     ],
 )
 def test_ops_general(op, targop):
diff --git a/pandas/tests/indexes/test_setops.py b/pandas/tests/indexes/test_setops.py
index 1dfeb4f2c6b5b..d126d32e627cd 100644
--- a/pandas/tests/indexes/test_setops.py
+++ b/pandas/tests/indexes/test_setops.py
@@ -320,8 +320,9 @@ def test_corner_union(self, index_flat, fname, sname, expected_name):
         # Test unions with various name combinations
         # Do not test MultiIndex or repeats
         if not index_flat.is_unique:
-            pytest.skip("Randomly generated index_flat was not unique.")
-        index = index_flat
+            index = index_flat.unique()
+        else:
+            index = index_flat
 
         # Test copy.union(copy)
         first = index.copy().set_names(fname)
@@ -363,8 +364,9 @@ def test_corner_union(self, index_flat, fname, sname, expected_name):
     )
     def test_union_unequal(self, index_flat, fname, sname, expected_name):
         if not index_flat.is_unique:
-            pytest.skip("Randomly generated index_flat was not unique.")
-        index = index_flat
+            index = index_flat.unique()
+        else:
+            index = index_flat
 
         # test copy.union(subset) - need sort for unicode and string
         first = index.copy().set_names(fname)
@@ -387,8 +389,9 @@ def test_corner_intersect(self, index_flat, fname, sname, expected_name):
         # GH#35847
         # Test intersections with various name combinations
         if not index_flat.is_unique:
-            pytest.skip("Randomly generated index_flat was not unique.")
-        index = index_flat
+            index = index_flat.unique()
+        else:
+            index = index_flat
 
         # Test copy.intersection(copy)
         first = index.copy().set_names(fname)
@@ -430,8 +433,9 @@ def test_corner_intersect(self, index_flat, fname, sname, expected_name):
     )
     def test_intersect_unequal(self, index_flat, fname, sname, expected_name):
         if not index_flat.is_unique:
-            pytest.skip("Randomly generated index_flat was not unique.")
-        index = index_flat
+            index = index_flat.unique()
+        else:
+            index = index_flat
 
         # test copy.intersection(subset) - need sort for unicode and string
         first = index.copy().set_names(fname)
diff --git a/pandas/tests/io/excel/test_writers.py b/pandas/tests/io/excel/test_writers.py
index 946c621ae2b6e..b9ea440b3e859 100644
--- a/pandas/tests/io/excel/test_writers.py
+++ b/pandas/tests/io/excel/test_writers.py
@@ -12,6 +12,7 @@
 import pytest
 
 from pandas.compat._constants import PY310
+from pandas.compat._optional import import_optional_dependency
 import pandas.util._test_decorators as td
 
 import pandas as pd
@@ -1309,7 +1310,9 @@ class TestExcelWriterEngineTests:
     def test_ExcelWriter_dispatch(self, klass, ext):
         with tm.ensure_clean(ext) as path:
             with ExcelWriter(path) as writer:
-                if ext == ".xlsx" and td.safe_import("xlsxwriter"):
+                if ext == ".xlsx" and bool(
+                    import_optional_dependency("xlsxwriter", errors="ignore")
+                ):
                     # xlsxwriter has preference over openpyxl if both installed
                     assert isinstance(writer, _XlsxWriter)
                 else:
diff --git a/pandas/tests/plotting/frame/test_frame_legend.py b/pandas/tests/plotting/frame/test_frame_legend.py
index d2924930667b6..402a4b9531e5d 100644
--- a/pandas/tests/plotting/frame/test_frame_legend.py
+++ b/pandas/tests/plotting/frame/test_frame_legend.py
@@ -231,7 +231,7 @@ def test_legend_name(self):
             "line",
             "bar",
             "barh",
-            pytest.param("kde", marks=td.skip_if_no_scipy),
+            pytest.param("kde", marks=td.skip_if_no("scipy")),
             "area",
             "hist",
         ],
diff --git a/pandas/tests/plotting/test_backend.py b/pandas/tests/plotting/test_backend.py
index c0ad8e0c9608d..8a8643415ae12 100644
--- a/pandas/tests/plotting/test_backend.py
+++ b/pandas/tests/plotting/test_backend.py
@@ -80,7 +80,7 @@ def test_setting_backend_without_plot_raises(monkeypatch):
     assert pandas.options.plotting.backend == "matplotlib"
 
 
-@td.skip_if_mpl
+@td.skip_if_installed("matplotlib")
 def test_no_matplotlib_ok():
     msg = (
         'matplotlib is required for plotting when the default backend "matplotlib" is '
diff --git a/pandas/tests/plotting/test_misc.py b/pandas/tests/plotting/test_misc.py
index 5b06cd1ab296a..84f9cd87db97c 100644
--- a/pandas/tests/plotting/test_misc.py
+++ b/pandas/tests/plotting/test_misc.py
@@ -30,7 +30,7 @@
 cm = pytest.importorskip("matplotlib.cm")
 
 
-@td.skip_if_mpl
+@td.skip_if_installed("matplotlib")
 def test_import_error_message():
     # GH-19810
     df = DataFrame({"A": [1, 2]})
diff --git a/pandas/tests/plotting/test_series.py b/pandas/tests/plotting/test_series.py
index 6719e39612f5d..f78f5c4879b9f 100644
--- a/pandas/tests/plotting/test_series.py
+++ b/pandas/tests/plotting/test_series.py
@@ -91,7 +91,7 @@ def test_plot_iseries(self, iseries):
             "line",
             "bar",
             "barh",
-            pytest.param("kde", marks=td.skip_if_no_scipy),
+            pytest.param("kde", marks=td.skip_if_no("scipy")),
             "hist",
             "box",
         ],
diff --git a/pandas/tests/series/methods/test_interpolate.py b/pandas/tests/series/methods/test_interpolate.py
index 426885bf41a1f..d854f0b787759 100644
--- a/pandas/tests/series/methods/test_interpolate.py
+++ b/pandas/tests/series/methods/test_interpolate.py
@@ -205,7 +205,7 @@ def test_interpolate_from_derivatives(self):
         [
             {},
             pytest.param(
-                {"method": "polynomial", "order": 1}, marks=td.skip_if_no_scipy
+                {"method": "polynomial", "order": 1}, marks=td.skip_if_no("scipy")
             ),
         ],
     )
@@ -253,7 +253,7 @@ def test_interpolate_non_ts(self):
         [
             {},
             pytest.param(
-                {"method": "polynomial", "order": 1}, marks=td.skip_if_no_scipy
+                {"method": "polynomial", "order": 1}, marks=td.skip_if_no("scipy")
             ),
         ],
     )
@@ -628,7 +628,7 @@ def test_interp_all_good(self):
         tm.assert_series_equal(result, s)
 
     @pytest.mark.parametrize(
-        "check_scipy", [False, pytest.param(True, marks=td.skip_if_no_scipy)]
+        "check_scipy", [False, pytest.param(True, marks=td.skip_if_no("scipy"))]
     )
     def test_interp_multiIndex(self, check_scipy):
         idx = MultiIndex.from_tuples([(0, "a"), (1, "b"), (2, "c")])
diff --git a/pandas/tests/strings/test_extract.py b/pandas/tests/strings/test_extract.py
index b8319e90e09a8..9ad9b1eca41d9 100644
--- a/pandas/tests/strings/test_extract.py
+++ b/pandas/tests/strings/test_extract.py
@@ -360,7 +360,7 @@ def test_extract_dataframe_capture_groups_index(index, any_string_dtype):
     data = ["A1", "B2", "C"]
 
     if len(index) < len(data):
-        pytest.skip("Index too short")
+        pytest.skip(f"Index needs more than {len(data)} values")
 
     index = index[: len(data)]
     s = Series(data, index=index, dtype=any_string_dtype)
diff --git a/pandas/tests/util/test_safe_import.py b/pandas/tests/util/test_safe_import.py
deleted file mode 100644
index bd07bea934ed3..0000000000000
--- a/pandas/tests/util/test_safe_import.py
+++ /dev/null
@@ -1,39 +0,0 @@
-import sys
-import types
-
-import pytest
-
-import pandas.util._test_decorators as td
-
-
-@pytest.mark.parametrize("name", ["foo", "hello123"])
-def test_safe_import_non_existent(name):
-    assert not td.safe_import(name)
-
-
-def test_safe_import_exists():
-    assert td.safe_import("pandas")
-
-
-@pytest.mark.parametrize("min_version,valid", [("0.0.0", True), ("99.99.99", False)])
-def test_safe_import_versions(min_version, valid):
-    result = td.safe_import("pandas", min_version=min_version)
-    result = result if valid else not result
-    assert result
-
-
-@pytest.mark.parametrize(
-    "min_version,valid", [(None, False), ("1.0", True), ("2.0", False)]
-)
-def test_safe_import_dummy(monkeypatch, min_version, valid):
-    mod_name = "hello123"
-
-    mod = types.ModuleType(mod_name)
-    mod.__version__ = "1.5"
-
-    if min_version is not None:
-        monkeypatch.setitem(sys.modules, mod_name, mod)
-
-    result = td.safe_import(mod_name, min_version=min_version)
-    result = result if valid else not result
-    assert result
diff --git a/pandas/tests/window/test_rolling_functions.py b/pandas/tests/window/test_rolling_functions.py
index 51f801ab3761b..5906ff52db098 100644
--- a/pandas/tests/window/test_rolling_functions.py
+++ b/pandas/tests/window/test_rolling_functions.py
@@ -346,7 +346,7 @@ def test_center_reindex_frame(frame, roll_func, kwargs, minp, fill_value):
         lambda x: x.rolling(window=10, min_periods=5).apply(sum, raw=True),
         pytest.param(
             lambda x: x.rolling(win_type="boxcar", window=10, min_periods=5).mean(),
-            marks=td.skip_if_no_scipy,
+            marks=td.skip_if_no("scipy"),
         ),
     ],
 )
@@ -508,7 +508,7 @@ def test_rolling_min_max_numeric_types(data_type):
         lambda x: x.rolling(window=10, min_periods=5).apply(sum, raw=True),
         pytest.param(
             lambda x: x.rolling(win_type="boxcar", window=10, min_periods=5).mean(),
-            marks=td.skip_if_no_scipy,
+            marks=td.skip_if_no("scipy"),
         ),
     ],
 )
diff --git a/pandas/util/_test_decorators.py b/pandas/util/_test_decorators.py
index e71b4881a58a4..2c1912bce856d 100644
--- a/pandas/util/_test_decorators.py
+++ b/pandas/util/_test_decorators.py
@@ -2,8 +2,8 @@
 This module provides decorator functions which can be applied to test objects
 in order to skip those objects when certain conditions occur. A sample use case
 is to detect if the platform is missing ``matplotlib``. If so, any test objects
-which require ``matplotlib`` and decorated with ``@td.skip_if_no_mpl`` will be
-skipped by ``pytest`` during the execution of the test suite.
+which require ``matplotlib`` and decorated with ``@td.skip_if_no("matplotlib")``
+will be skipped by ``pytest`` during the execution of the test suite.
 
 To illustrate, after importing this module:
 
@@ -11,13 +11,13 @@
 
 The decorators can be applied to classes:
 
-@td.skip_if_some_reason
+@td.skip_if_no("package")
 class Foo:
     ...
 
 Or individual functions:
 
-@td.skip_if_some_reason
+@td.skip_if_no("package")
 def test_foo():
     ...
 
@@ -44,59 +44,7 @@ def test_foo():
     IS64,
     is_platform_windows,
 )
-
-from pandas.core.computation.expressions import (
-    NUMEXPR_INSTALLED,
-    USE_NUMEXPR,
-)
-from pandas.util.version import Version
-
-
-def safe_import(mod_name: str, min_version: str | None = None):
-    """
-    Parameters
-    ----------
-    mod_name : str
-        Name of the module to be imported
-    min_version : str, default None
-        Minimum required version of the specified mod_name
-
-    Returns
-    -------
-    object
-        The imported module if successful, or False
-    """
-    try:
-        mod = __import__(mod_name)
-    except ImportError:
-        return False
-
-    if not min_version:
-        return mod
-    else:
-        import sys
-
-        version = getattr(sys.modules[mod_name], "__version__")
-        if version and Version(version) >= Version(min_version):
-            return mod
-
-    return False
-
-
-def _skip_if_not_us_locale() -> bool:
-    lang, _ = locale.getlocale()
-    if lang != "en_US":
-        return True
-    return False
-
-
-def _skip_if_no_scipy() -> bool:
-    return not (
-        safe_import("scipy.stats")
-        and safe_import("scipy.sparse")
-        and safe_import("scipy.interpolate")
-        and safe_import("scipy.signal")
-    )
+from pandas.compat._optional import import_optional_dependency
 
 
 def skip_if_installed(package: str) -> pytest.MarkDecorator:
@@ -115,7 +63,8 @@ def skip_if_installed(package: str) -> pytest.MarkDecorator:
         parametrization mark.
     """
     return pytest.mark.skipif(
-        safe_import(package), reason=f"Skipping because {package} is installed."
+        bool(import_optional_dependency(package, errors="ignore")),
+        reason=f"Skipping because {package} is installed.",
     )
 
 
@@ -154,25 +103,20 @@ def skip_if_no(package: str, min_version: str | None = None) -> pytest.MarkDecor
     if min_version:
         msg += f" satisfying a min_version of {min_version}"
     return pytest.mark.skipif(
-        not safe_import(package, min_version=min_version), reason=msg
+        not bool(
+            import_optional_dependency(
+                package, errors="ignore", min_version=min_version
+            )
+        ),
+        reason=msg,
     )
 
 
-skip_if_mpl = pytest.mark.skipif(
-    bool(safe_import("matplotlib")), reason="matplotlib is present"
-)
 skip_if_32bit = pytest.mark.skipif(not IS64, reason="skipping for 32 bit")
 skip_if_windows = pytest.mark.skipif(is_platform_windows(), reason="Running on Windows")
 skip_if_not_us_locale = pytest.mark.skipif(
-    _skip_if_not_us_locale(),
-    reason=f"Specific locale is set {locale.getlocale()[0]}",
-)
-skip_if_no_scipy = pytest.mark.skipif(
-    _skip_if_no_scipy(), reason="Missing SciPy requirement"
-)
-skip_if_no_ne = pytest.mark.skipif(
-    not USE_NUMEXPR,
-    reason=f"numexpr enabled->{USE_NUMEXPR}, installed->{NUMEXPR_INSTALLED}",
+    locale.getlocale()[0] != "en_US",
+    reason=f"Set local {locale.getlocale()[0]} is not en_US",
 )