diff --git a/Doc/library/multiprocessing.shared_memory.rst b/Doc/library/multiprocessing.shared_memory.rst
index f453e6403d932d..671130d9b29fc0 100644
--- a/Doc/library/multiprocessing.shared_memory.rst
+++ b/Doc/library/multiprocessing.shared_memory.rst
@@ -36,7 +36,7 @@ or other communications requiring the serialization/deserialization and
 copying of data.
 
 
-.. class:: SharedMemory(name=None, create=False, size=0)
+.. class:: SharedMemory(name=None, create=False, size=0, *, track=True)
 
    Creates a new shared memory block or attaches to an existing shared
    memory block.  Each shared memory block is assigned a unique name.
@@ -64,26 +64,45 @@ copying of data.
    memory block may be larger or equal to the size requested.  When attaching
    to an existing shared memory block, the ``size`` parameter is ignored.
 
+   *track*, when enabled, registers the shared memory block with a resource
+   tracker process on platforms where the OS does not do this automatically.
+   The resource tracker ensures proper cleanup of the shared memory even
+   if all other processes with access to the memory exit without doing so.
+   Python processes created from a common ancestor using :mod:`multiprocessing`
+   facilities share a single resource tracker process, and the lifetime of
+   shared memory segments is handled automatically among these processes.
+   Python processes created in any other way will receive their own
+   resource tracker when accessing shared memory with *track* enabled.
+   This will cause the shared memory to be deleted by the resource tracker
+   of the first process that terminates.
+   To avoid this issue, users of :mod:`subprocess` or standalone Python
+   processes should set *track* to ``False`` when there is already another
+   process in place that does the bookkeeping.
+   *track* is ignored on Windows, which has its own tracking and
+   automatically deletes shared memory when all handles to it have been closed.
+
+   .. versionchanged:: 3.13 Added *track* parameter.
+
    .. method:: close()
 
-      Closes access to the shared memory from this instance.  In order to
-      ensure proper cleanup of resources, all instances should call
-      ``close()`` once the instance is no longer needed.  Note that calling
-      ``close()`` does not cause the shared memory block itself to be
-      destroyed.
+      Closes the file descriptor/handle to the shared memory from this
+      instance.  :meth:`close()` should be called once access to the shared
+      memory block from this instance is no longer needed.  Depending
+      on operating system, the underlying memory may or may not be freed
+      even if all handles to it have been closed.  To ensure proper cleanup,
+      use the :meth:`unlink()` method.
 
    .. method:: unlink()
 
-      Requests that the underlying shared memory block be destroyed.  In
-      order to ensure proper cleanup of resources, ``unlink()`` should be
-      called once (and only once) across all processes which have need
-      for the shared memory block.  After requesting its destruction, a
-      shared memory block may or may not be immediately destroyed and
-      this behavior may differ across platforms.  Attempts to access data
-      inside the shared memory block after ``unlink()`` has been called may
-      result in memory access errors.  Note: the last process relinquishing
-      its hold on a shared memory block may call ``unlink()`` and
-      :meth:`close()` in either order.
+      Deletes the underlying shared memory block.  This should be called only
+      once per shared memory block regardless of the number of handles to it,
+      even in other processes.
+      :meth:`unlink()` and :meth:`close()` can be called in any order, but
+      trying to access data inside a shared memory block after :meth:`unlink()`
+      may result in memory access errors, depending on platform.
+
+      This method has no effect on Windows, where the only way to delete a
+      shared memory block is to close all handles.
 
    .. attribute:: buf
 
diff --git a/Lib/multiprocessing/shared_memory.py b/Lib/multiprocessing/shared_memory.py
index 9a1e5aa17b87a2..67e70fdc27cf31 100644
--- a/Lib/multiprocessing/shared_memory.py
+++ b/Lib/multiprocessing/shared_memory.py
@@ -71,8 +71,9 @@ class SharedMemory:
     _flags = os.O_RDWR
     _mode = 0o600
     _prepend_leading_slash = True if _USE_POSIX else False
+    _track = True
 
-    def __init__(self, name=None, create=False, size=0):
+    def __init__(self, name=None, create=False, size=0, *, track=True):
         if not size >= 0:
             raise ValueError("'size' must be a positive integer")
         if create:
@@ -82,6 +83,7 @@ def __init__(self, name=None, create=False, size=0):
         if name is None and not self._flags & os.O_EXCL:
             raise ValueError("'name' can only be None if create=True")
 
+        self._track = track
         if _USE_POSIX:
 
             # POSIX Shared Memory
@@ -116,8 +118,8 @@ def __init__(self, name=None, create=False, size=0):
             except OSError:
                 self.unlink()
                 raise
-
-            resource_tracker.register(self._name, "shared_memory")
+            if self._track:
+                resource_tracker.register(self._name, "shared_memory")
 
         else:
 
@@ -236,12 +238,20 @@ def close(self):
     def unlink(self):
         """Requests that the underlying shared memory block be destroyed.
 
-        In order to ensure proper cleanup of resources, unlink should be
-        called once (and only once) across all processes which have access
-        to the shared memory block."""
+        Unlink should be called once (and only once) across all handles
+        which have access to the shared memory block, even if these
+        handles belong to different processes. Closing and unlinking may
+        happen in any order, but trying to access data inside a shared
+        memory block after unlinking may result in memory errors,
+        depending on platform.
+
+        This method has no effect on Windows, where the only way to
+        delete a shared memory block is to close all handles."""
+
         if _USE_POSIX and self._name:
             _posixshmem.shm_unlink(self._name)
-            resource_tracker.unregister(self._name, "shared_memory")
+            if self._track:
+                resource_tracker.unregister(self._name, "shared_memory")
 
 
 _encoding = "utf8"
diff --git a/Lib/test/_test_multiprocessing.py b/Lib/test/_test_multiprocessing.py
index ec003d8dc4314d..a94eb6c0ae4b8e 100644
--- a/Lib/test/_test_multiprocessing.py
+++ b/Lib/test/_test_multiprocessing.py
@@ -4455,6 +4455,59 @@ def test_shared_memory_cleaned_after_process_termination(self):
                     "resource_tracker: There appear to be 1 leaked "
                     "shared_memory objects to clean up at shutdown", err)
 
+    @unittest.skipIf(os.name != "posix", "resource_tracker is posix only")
+    def test_shared_memory_untracking(self):
+        # gh-82300: When a separate Python process accesses shared memory
+        # with track=False, it must not cause the memory to be deleted
+        # when terminating.
+        cmd = '''if 1:
+            import sys
+            from multiprocessing.shared_memory import SharedMemory
+            mem = SharedMemory(create=False, name=sys.argv[1], track=False)
+            mem.close()
+        '''
+        mem = shared_memory.SharedMemory(create=True, size=10)
+        # The resource tracker shares pipes with the subprocess, and so
+        # err existing means that the tracker process has terminated now.
+        try:
+            rc, out, err = script_helper.assert_python_ok("-c", cmd, mem.name)
+            self.assertNotIn(b"resource_tracker", err)
+            self.assertEqual(rc, 0)
+            mem2 = shared_memory.SharedMemory(create=False, name=mem.name)
+            mem2.close()
+        finally:
+            try:
+                mem.unlink()
+            except OSError:
+                pass
+            mem.close()
+
+    @unittest.skipIf(os.name != "posix", "resource_tracker is posix only")
+    def test_shared_memory_tracking(self):
+        # gh-82300: When a separate Python process accesses shared memory
+        # with track=True, it must cause the memory to be deleted when
+        # terminating.
+        cmd = '''if 1:
+            import sys
+            from multiprocessing.shared_memory import SharedMemory
+            mem = SharedMemory(create=False, name=sys.argv[1], track=True)
+            mem.close()
+        '''
+        mem = shared_memory.SharedMemory(create=True, size=10)
+        try:
+            rc, out, err = script_helper.assert_python_ok("-c", cmd, mem.name)
+            self.assertEqual(rc, 0)
+            self.assertIn(
+                b"resource_tracker: There appear to be 1 leaked "
+                b"shared_memory objects to clean up at shutdown", err)
+        finally:
+            try:
+                mem.unlink()
+            except OSError:
+                pass
+            resource_tracker.unregister(mem._name, "shared_memory")
+            mem.close()
+
 #
 # Test to verify that `Finalize` works.
 #
diff --git a/Misc/NEWS.d/next/Library/2023-10-12-18-19-47.gh-issue-82300.P8-O38.rst b/Misc/NEWS.d/next/Library/2023-10-12-18-19-47.gh-issue-82300.P8-O38.rst
new file mode 100644
index 00000000000000..d7e6b225489b99
--- /dev/null
+++ b/Misc/NEWS.d/next/Library/2023-10-12-18-19-47.gh-issue-82300.P8-O38.rst
@@ -0,0 +1 @@
+Add ``track`` parameter to :class:`multiprocessing.shared_memory.SharedMemory` that allows using shared memory blocks without having to register with the POSIX resource tracker that automatically releases them upon process exit.