diff --git a/awswrangler/_data_types.py b/awswrangler/_data_types.py
index 432b3f72e..3b54640c9 100644
--- a/awswrangler/_data_types.py
+++ b/awswrangler/_data_types.py
@@ -344,7 +344,7 @@ def athena2pyarrow(dtype: str, df_type: str | None = None) -> pa.DataType:  # no
             return pa.timestamp(unit="ns")
     if dtype == "date":
         return pa.date32()
-    if dtype in ("binary" or "varbinary"):
+    if dtype in ("binary", "varbinary"):
         return pa.binary()
     if dtype.startswith("decimal") is True:
         precision, scale = dtype.replace("decimal(", "").replace(")", "").split(sep=",")
diff --git a/tests/unit/test_data_types.py b/tests/unit/test_data_types.py
new file mode 100644
index 000000000..233ad39fc
--- /dev/null
+++ b/tests/unit/test_data_types.py
@@ -0,0 +1,29 @@
+import pyarrow as pa
+import pytest
+
+from awswrangler._data_types import athena2pandas, athena2pyarrow
+from awswrangler.exceptions import UnsupportedType
+
+
+@pytest.mark.parametrize(
+    "dtype,expected",
+    [
+        ("binary", pa.binary()),
+        ("varbinary", pa.binary()),
+        ("BINARY", pa.binary()),
+        ("VARBINARY", pa.binary()),
+    ],
+)
+def test_athena2pyarrow_binary_types(dtype, expected):
+    assert athena2pyarrow(dtype) == expected
+
+
+@pytest.mark.parametrize("dtype", ["i", "n", "ary", "bin"])
+def test_athena2pyarrow_rejects_binary_substrings(dtype):
+    with pytest.raises(UnsupportedType, match=f"Unsupported Athena type: {dtype}"):
+        athena2pyarrow(dtype)
+
+
+@pytest.mark.parametrize("dtype", ["binary", "varbinary"])
+def test_athena2pandas_binary_types(dtype):
+    assert athena2pandas(dtype) == "bytes"