diff --git a/doc/whats_new/v1.1.rst b/doc/whats_new/v1.1.rst
index 87ffc36d342fa..93300bf67d4a5 100644
--- a/doc/whats_new/v1.1.rst
+++ b/doc/whats_new/v1.1.rst
@@ -38,7 +38,6 @@ Changelog
     :pr:`123456` by :user:`Joe Bloggs <joeongithub>`.
     where 123456 is the *pull request* number, not the issue number.
 
-
 :mod:`sklearn.calibration`
 ..........................
 
@@ -46,6 +45,14 @@ Changelog
   `pos_label` to specify the positive class label.
   :pr:`21032` by :user:`Guillaume Lemaitre <glemaitre>`.
 
+:mod:`sklearn.ensemble`
+...........................
+
+- |Fix| Fixed a bug that could produce a segfault in rare cases for
+  :class:`ensemble.HistGradientBoostingClassifier` and
+  :class:`ensemble.HistGradientBoostingRegressor`.
+  :pr:`21130` :user:`Christian Lorentzen <lorentzenchr>`.
+
 :mod:`sklearn.linear_model`
 ...........................
 
diff --git a/sklearn/ensemble/_hist_gradient_boosting/splitting.pyx b/sklearn/ensemble/_hist_gradient_boosting/splitting.pyx
index 08ae7aaf0862c..232cf094876cb 100644
--- a/sklearn/ensemble/_hist_gradient_boosting/splitting.pyx
+++ b/sklearn/ensemble/_hist_gradient_boosting/splitting.pyx
@@ -388,11 +388,25 @@ cdef class Splitter:
                     &left_indices_buffer[offset_in_buffers[thread_idx]],
                     sizeof(unsigned int) * left_counts[thread_idx]
                 )
-                memcpy(
-                    &sample_indices[right_offset[thread_idx]],
-                    &right_indices_buffer[offset_in_buffers[thread_idx]],
-                    sizeof(unsigned int) * right_counts[thread_idx]
-                )
+                if right_counts[thread_idx] > 0:
+                    # If we're splitting the rightmost node of the tree, i.e. the
+                    # rightmost node in the partition array, and if n_threads >= 2, one
+                    # might have right_counts[-1] = 0 and right_offset[-1] = len(sample_indices)
+                    # leading to evaluating
+                    #
+                    #    &sample_indices[right_offset[-1]] = &samples_indices[n_samples_at_node]
+                    #                                      = &partition[n_samples_in_tree]
+                    #
+                    # which is an out-of-bounds read access that can cause a segmentation fault.
+                    # When boundscheck=True, removing this check produces this exception:
+                    #
+                    #    IndexError: Out of bounds on buffer access
+                    #
+                    memcpy(
+                        &sample_indices[right_offset[thread_idx]],
+                        &right_indices_buffer[offset_in_buffers[thread_idx]],
+                        sizeof(unsigned int) * right_counts[thread_idx]
+                    )
 
         return (sample_indices[:right_child_position],
                 sample_indices[right_child_position:],
@@ -839,7 +853,7 @@ cdef class Splitter:
         # other category. The low-support categories will always be mapped to
         # the right child. We scan the sorted categories array from left to
         # right and from right to left, and we stop at the middle.
-        
+
         # Considering ordered categories A B C D, with E being a low-support
         # category: A B C D
         #              ^