diff --git a/pandas/core/arrays/arrow/array.py b/pandas/core/arrays/arrow/array.py
index d4f0e4681405b..8878a20bb5b42 100644
--- a/pandas/core/arrays/arrow/array.py
+++ b/pandas/core/arrays/arrow/array.py
@@ -439,6 +439,10 @@ def _cast_pointwise_result(self, values) -> ArrayLike:
             #  or test_agg_lambda_complex128_dtype_conversion for complex values
             return super()._cast_pointwise_result(values)
 
+        if pa.types.is_null(arr.type):
+            if lib.infer_dtype(values) == "decimal":
+                # GH#62522; the specific decimal precision here is arbitrary
+                arr = arr.cast(pa.decimal32(1))
         if pa.types.is_duration(arr.type):
             # workaround for https://github.com/apache/arrow/issues/40620
             result = ArrowExtensionArray._from_sequence(values)
diff --git a/pandas/tests/extension/test_arrow.py b/pandas/tests/extension/test_arrow.py
index c810f098f15cf..f39f7acdc57bf 100644
--- a/pandas/tests/extension/test_arrow.py
+++ b/pandas/tests/extension/test_arrow.py
@@ -3700,3 +3700,15 @@ def test_pow_with_all_na_float():
     result = s.pow(2)
     expected = pd.Series([pd.NA, pd.NA], dtype="float64[pyarrow]")
     tm.assert_series_equal(result, expected)
+
+
+def test_cast_pontwise_result_decimal_nan():
+    # GH#62522 we don't want to get back null[pyarrow] here
+    ser = pd.Series([], dtype="float64[pyarrow]")
+    arr = ser.array
+    item = Decimal("NaN")
+
+    result = arr._cast_pointwise_result([item])
+
+    pa_type = result.dtype.pyarrow_dtype
+    assert pa.types.is_decimal(pa_type)