diff --git a/test/xpu/run_test_with_skip_arc.py b/test/xpu/run_test_with_skip_arc.py
index 4609cd6e5..c70256dd6 100644
--- a/test/xpu/run_test_with_skip_arc.py
+++ b/test/xpu/run_test_with_skip_arc.py
@@ -2,13 +2,19 @@
 import sys
 from skip_list_common import skip_dict
 from skip_list_arc import skip_dict as skip_dict_specifical
+from skip_list_win import skip_dict as skip_dict_win
 from xpu_test_utils import launch_test
 
 
 res = 0
+IS_WINDOWS = sys.platform == "win32"
 
 for key in skip_dict:
-    skip_list = skip_dict[key] if key not in skip_dict_specifical else skip_dict[key] + skip_dict_specifical[key]
+    skip_list = skip_dict[key]
+    if key in skip_dict_specifical:
+        skip_list += skip_dict_specifical[key]
+    if IS_WINDOWS and key in skip_dict_win:
+        skip_list += skip_dict_win[key]
     res += launch_test(key, skip_list)
 
 exit_code = os.WEXITSTATUS(res)
diff --git a/test/xpu/run_test_with_skip_mtl.py b/test/xpu/run_test_with_skip_mtl.py
index b53c29bf7..828466b11 100644
--- a/test/xpu/run_test_with_skip_mtl.py
+++ b/test/xpu/run_test_with_skip_mtl.py
@@ -2,12 +2,18 @@
 import sys
 from skip_list_common import skip_dict
 from skip_list_mtl import skip_dict as skip_dict_specifical
+from skip_list_win import skip_dict as skip_dict_win
 from xpu_test_utils import launch_test
 
 res = 0
+IS_WINDOWS = sys.platform == "win32"
 
 for key in skip_dict:
-    skip_list = skip_dict[key] if key not in skip_dict_specifical else skip_dict[key] + skip_dict_specifical[key]
+    skip_list = skip_dict[key]
+    if key in skip_dict_specifical:
+        skip_list += skip_dict_specifical[key]
+    if IS_WINDOWS and key in skip_dict_win:
+        skip_list += skip_dict_win[key]
     res += launch_test(key, skip_list)
 
 exit_code = os.WEXITSTATUS(res)
diff --git a/test/xpu/skip_list_common.py b/test/xpu/skip_list_common.py
index 52a96e3d7..dbb757686 100644
--- a/test/xpu/skip_list_common.py
+++ b/test/xpu/skip_list_common.py
@@ -749,7 +749,7 @@
         "test_gather_backward_with_empty_index_tensor_sparse_grad_True_xpu_float64",
     ),
 
-    "test_autograd_fallback.py": None,
+    "test_autograd_fallback_xpu.py": None,
 
     "test_sort_and_select_xpu.py": ("test_sort_large_slice_xpu",),  # Hard code CUDA
 
diff --git a/test/xpu/skip_list_win.py b/test/xpu/skip_list_win.py
new file mode 100644
index 000000000..35bf3014a
--- /dev/null
+++ b/test/xpu/skip_list_win.py
@@ -0,0 +1,75 @@
+skip_dict = {
+    # SYCL default context is not supported on Windows
+    "test_tensor_creation_ops_xpu.py": (
+        "test_alias_from_dlpack_xpu_bfloat16",
+        "test_alias_from_dlpack_xpu_complex128",
+        "test_alias_from_dlpack_xpu_complex64",
+        "test_alias_from_dlpack_xpu_float16",
+        "test_alias_from_dlpack_xpu_float32",
+        "test_alias_from_dlpack_xpu_float64",
+        "test_alias_from_dlpack_xpu_int16",
+        "test_alias_from_dlpack_xpu_int32",
+        "test_alias_from_dlpack_xpu_int64",
+        "test_alias_from_dlpack_xpu_int8",
+        "test_alias_from_dlpack_xpu_uint8",
+        "test_copy_from_dlpack_xpu_bfloat16",
+        "test_copy_from_dlpack_xpu_complex128",
+        "test_copy_from_dlpack_xpu_complex64",
+        "test_copy_from_dlpack_xpu_float16",
+        "test_copy_from_dlpack_xpu_float32",
+        "test_copy_from_dlpack_xpu_float64",
+        "test_copy_from_dlpack_xpu_int16",
+        "test_copy_from_dlpack_xpu_int32",
+        "test_copy_from_dlpack_xpu_int64",
+        "test_copy_from_dlpack_xpu_int8",
+        "test_copy_from_dlpack_xpu_uint8"
+    ),
+    # SYCL compiler issue where host and device results differ for math ops with complex dtypes
+    "test_unary_ufuncs_xpu.py": (
+        "test_reference_numerics_extremal__refs_atanh_xpu_complex128",
+        "test_reference_numerics_extremal__refs_atanh_xpu_complex64",
+        "test_reference_numerics_extremal__refs_nn_functional_tanhshrink_xpu_complex128",
+        "test_reference_numerics_extremal__refs_sin_xpu_complex128",
+        "test_reference_numerics_extremal__refs_sin_xpu_complex64",
+        "test_reference_numerics_extremal__refs_sinh_xpu_complex128",
+        "test_reference_numerics_extremal__refs_sinh_xpu_complex64",
+        "test_reference_numerics_extremal__refs_tan_xpu_complex128",
+        "test_reference_numerics_extremal__refs_tan_xpu_complex64",
+        "test_reference_numerics_extremal_atanh_xpu_complex128",
+        "test_reference_numerics_extremal_atanh_xpu_complex64",
+        "test_reference_numerics_extremal_nn_functional_tanhshrink_xpu_complex128",
+        "test_reference_numerics_extremal_sin_xpu_complex128",
+        "test_reference_numerics_extremal_sin_xpu_complex64",
+        "test_reference_numerics_extremal_sinh_xpu_complex128",
+        "test_reference_numerics_extremal_sinh_xpu_complex64",
+        "test_reference_numerics_extremal_square_xpu_complex128",
+        "test_reference_numerics_extremal_square_xpu_complex64",
+        "test_reference_numerics_extremal_tan_xpu_complex128",
+        "test_reference_numerics_extremal_tan_xpu_complex64",
+        "test_reference_numerics_large__refs_cos_xpu_complex128",
+        "test_reference_numerics_large__refs_cos_xpu_complex32",
+        "test_reference_numerics_large__refs_cos_xpu_complex64",
+        "test_reference_numerics_large__refs_cosh_xpu_complex32",
+        "test_reference_numerics_large__refs_exp_xpu_complex128",
+        "test_reference_numerics_large__refs_exp_xpu_complex32",
+        "test_reference_numerics_large__refs_exp_xpu_complex64",
+        "test_reference_numerics_large__refs_sin_xpu_complex128",
+        "test_reference_numerics_large__refs_sin_xpu_complex32",
+        "test_reference_numerics_large__refs_sin_xpu_complex64",
+        "test_reference_numerics_large__refs_sinh_xpu_complex32",
+        "test_reference_numerics_large__refs_tan_xpu_complex32",
+        "test_reference_numerics_large_cos_xpu_complex128",
+        "test_reference_numerics_large_cos_xpu_complex32",
+        "test_reference_numerics_large_cos_xpu_complex64",
+        "test_reference_numerics_large_exp_xpu_complex128",
+        "test_reference_numerics_large_exp_xpu_complex64",
+        "test_reference_numerics_large_sin_xpu_complex128",
+        "test_reference_numerics_large_sin_xpu_complex32",
+        "test_reference_numerics_large_sin_xpu_complex64",
+        "test_reference_numerics_small_acos_xpu_complex32",
+    ),
+    "test_ops_xpu.py": (
+        "test_compare_cpu_pow_xpu_bfloat16",  # https://github.com/intel/torch-xpu-ops/pull/764
+    ),
+
+}
diff --git a/test/xpu/test_autograd_fallback.py b/test/xpu/test_autograd_fallback_xpu.py
similarity index 100%
rename from test/xpu/test_autograd_fallback.py
rename to test/xpu/test_autograd_fallback_xpu.py