diff --git a/doc/source/whatsnew/v2.1.2.rst b/doc/source/whatsnew/v2.1.2.rst
index db65e0b9f4315..a608cd136c559 100644
--- a/doc/source/whatsnew/v2.1.2.rst
+++ b/doc/source/whatsnew/v2.1.2.rst
@@ -17,6 +17,7 @@ Fixed regressions
 - Fixed regression in :meth:`DataFrame.resample` which was extrapolating back to ``origin`` when ``origin`` was outside its bounds (:issue:`55064`)
 - Fixed regression in :meth:`DataFrame.sort_index` which was not sorting correctly when the index was a sliced :class:`MultiIndex` (:issue:`55379`)
 - Fixed performance regression with wide DataFrames, typically involving methods where all columns were accessed individually (:issue:`55256`, :issue:`55245`)
+- Fixed regression in :func:`merge_asof` raising ``TypeError`` for ``by`` with datetime and timedelta dtypes (:issue:`55453`)
 
 .. ---------------------------------------------------------------------------
 .. _whatsnew_212.bug_fixes:
diff --git a/pandas/core/reshape/merge.py b/pandas/core/reshape/merge.py
index 9cdf513ad7da3..8d826ff6ae0d8 100644
--- a/pandas/core/reshape/merge.py
+++ b/pandas/core/reshape/merge.py
@@ -2167,6 +2167,10 @@ def _get_join_indexers(self) -> tuple[npt.NDArray[np.intp], npt.NDArray[np.intp]
                 # TODO: conversions for EAs that can be no-copy.
                 lbv = np.asarray(lbv)
                 rbv = np.asarray(rbv)
+                if needs_i8_conversion(lbv.dtype):
+                    lbv = lbv.view("i8")
+                if needs_i8_conversion(rbv.dtype):
+                    rbv = rbv.view("i8")
             else:
                 # We get here with non-ndarrays in test_merge_by_col_tz_aware
                 #  and test_merge_groupby_multiple_column_with_categorical_column
diff --git a/pandas/tests/reshape/merge/test_merge_asof.py b/pandas/tests/reshape/merge/test_merge_asof.py
index 8ada42898f947..d7198aa0e3500 100644
--- a/pandas/tests/reshape/merge/test_merge_asof.py
+++ b/pandas/tests/reshape/merge/test_merge_asof.py
@@ -1345,6 +1345,34 @@ def test_by_mixed_tz_aware(self):
         expected["value_y"] = np.array([np.nan], dtype=object)
         tm.assert_frame_equal(result, expected)
 
+    @pytest.mark.parametrize("dtype", ["m8[ns]", "M8[us]"])
+    def test_by_datelike(self, dtype):
+        # GH 55453
+        left = pd.DataFrame(
+            {
+                "by_col": np.array([1], dtype=dtype),
+                "on_col": [2],
+                "value": ["a"],
+            }
+        )
+        right = pd.DataFrame(
+            {
+                "by_col": np.array([1], dtype=dtype),
+                "on_col": [1],
+                "value": ["b"],
+            }
+        )
+        result = merge_asof(left, right, by="by_col", on="on_col")
+        expected = pd.DataFrame(
+            {
+                "by_col": np.array([1], dtype=dtype),
+                "on_col": [2],
+                "value_x": ["a"],
+                "value_y": ["b"],
+            }
+        )
+        tm.assert_frame_equal(result, expected)
+
     def test_timedelta_tolerance_nearest(self, unit):
         # GH 27642
         if unit == "s":