diff --git a/keystonemiddleware/auth_token/_cache.py b/keystonemiddleware/auth_token/_cache.py
index c52377ab..9cd5b00b 100644
--- a/keystonemiddleware/auth_token/_cache.py
+++ b/keystonemiddleware/auth_token/_cache.py
@@ -19,7 +19,7 @@ import six
 from keystonemiddleware.auth_token import _exceptions as exc
 from keystonemiddleware.auth_token import _memcache_crypt as memcache_crypt
 from keystonemiddleware.auth_token import _memcache_pool as memcache_pool
-from keystonemiddleware.i18n import _, _LE
+from keystonemiddleware.i18n import _, _LE, _LW
 from keystonemiddleware.openstack.common import memorycache
 
 
@@ -54,8 +54,18 @@ class _EnvCachePool(object):
 class _CachePool(list):
     """A lazy pool of cache references."""
 
-    def __init__(self, memcached_servers):
+    def __init__(self, memcached_servers, log):
         self._memcached_servers = memcached_servers
+        if not self._memcached_servers:
+            log.warning(_LW(
+                "Using the in-process token cache is deprecated as of the "
+                "4.2.0 release and may be removed in the 5.0.0 release or "
+                "the 'O' development cycle. The in-process cache causes "
+                "inconsistent results and high memory usage. When the feature "
+                "is removed the auth_token middleware will not cache tokens "
+                "by default which may result in performance issues. It is "
+                "recommended to use  memcache for the auth_token token cache "
+                "by setting the memcached_servers option."))
 
     @contextlib.contextmanager
     def reserve(self):
@@ -125,7 +135,7 @@ class TokenCache(object):
                                        **self._memcache_pool_options)
 
         else:
-            return _CachePool(self._memcached_servers)
+            return _CachePool(self._memcached_servers, self._LOG)
 
     def initialize(self, env):
         if self._initialized:
diff --git a/releasenotes/notes/deprecate-caching-tokens-in-process-a412b0f1dea84cb9.yaml b/releasenotes/notes/deprecate-caching-tokens-in-process-a412b0f1dea84cb9.yaml
new file mode 100644
index 00000000..6712ffe2
--- /dev/null
+++ b/releasenotes/notes/deprecate-caching-tokens-in-process-a412b0f1dea84cb9.yaml
@@ -0,0 +1,18 @@
+---
+deprecations:
+  - >
+    With the release of 4.2.0 of keystonemiddleware we no longer recommend
+    using the in-process token cache. In-process caching may result in
+    inconsistent validation, poor UX and race conditions.
+
+    It is recommended that the `memcached_servers` option is set in the
+    `keystone_authtoken` configuration section of the various services (e.g.
+    nova, glance, ...) with the endpoint of running memcached server(s).
+
+    When the feature is removed, not setting the `memcached_servers`
+    option will cause keystone to validate tokens more frequently, increasing
+    load. In production, use of caching is highly recommended.
+
+    This feature is deprecated as of 4.2.0 and is targeted for removal in
+    keystonemiddleware 5.0.0 or in the `O` development cycle, whichever is
+    later.