diff --git a/holoviews/core/util/__init__.py b/holoviews/core/util/__init__.py
index 906548f80a..ca9779a73f 100644
--- a/holoviews/core/util/__init__.py
+++ b/holoviews/core/util/__init__.py
@@ -2255,7 +2255,7 @@ def validate_regular_sampling(values, rtol=10e-6):
     if len(diffs) < 1:
         return True
     min_diff = diffs.min()
-    return abs(min_diff - diffs.max()) < abs(min_diff * rtol)
+    return abs(min_diff - diffs.max()) <= abs(min_diff * rtol)
 
 
 def compute_density(start, end, length, time_unit="us"):
diff --git a/holoviews/tests/element/test_image.py b/holoviews/tests/element/test_image.py
index 74aaca1463..506660fb25 100644
--- a/holoviews/tests/element/test_image.py
+++ b/holoviews/tests/element/test_image.py
@@ -4,6 +4,8 @@
 
 from __future__ import annotations
 
+import datetime as dt
+
 import numpy as np
 import pytest
 
@@ -74,6 +76,55 @@ def test_image_rtol_config(self):
         hv.Image({"vals": vals, "xs": xs, "ys": ys}, ["xs", "ys"], "vals")
         hv.config.image_rtol = image_rtol
 
+    @pytest.mark.parametrize(
+        "x",
+        [
+            # max(|step|) / min(|step|) - 1 = 0.0015
+            [dt.datetime(2017, 1, 1, 0, 0, 0, ms) for ms in [0, 10000, 20015]],
+            *(
+                np.array([0, 10000, 20015]).astype(f"datetime64[{unit}]")
+                for unit in ["D", "h", "m", "s", "ms", "us"]
+            ),
+        ],
+    )
+    def test_image_datetime_rtol_failure(self, x):
+        """
+        Constructing an Image with a date-time coordinate warns about irregular sampling
+        if, approximately speaking, one less than the ratio of the max and min steps
+        exceeds the default `image_rtol`.
+        """
+        y = np.arange(2)
+        hv.Image((x, y, np.zeros((len(y), len(x)))))
+        self.log_handler.assert_endswith(
+            "WARNING",
+            "set a higher tolerance on hv.config.image_rtol or the rtol parameter in "
+            "the Image constructor.",
+        )
+
+    @pytest.mark.parametrize(
+        "x",
+        [
+            # |granularity / step| = 1/400 = 0.0025
+            [dt.datetime(2017, 1, 1, 0, 0, 0, ms) for ms in [0, 400]],
+            # |granularity / step| = 1/800 = 0.00125
+            *(
+                np.array([10000, 10800]).astype(f"datetime64[{unit}]")
+                for unit in ["D", "h", "m", "s", "ms", "us"]
+            ),
+        ],
+    )
+    def test_image_datetime_rtol_granularity(self, x):
+        """
+        Constructing an Image with an evenly spaced date-time coordinate does not warn
+        about irregular sampling, even when the ratio of the date-time granularity to
+        the coordinate step is greater than the default `image_rtol` (or twice
+        `image_rtol` for `dt.datetime` types because of how `dt.timedelta` instances
+        round when multiplied by a `float`).
+        """
+        y = np.arange(2)
+        hv.Image((x, y, np.zeros((len(y), len(x)))))
+        assert len(self.log_handler.tail("WARNING", n=1)) == 0
+
     def test_image_clone(self):
         vals = np.random.rand(20, 20)
         xs = np.linspace(0, 10, 20)