diff --git a/changelog.d/13632.bugfix b/changelog.d/13632.bugfix
new file mode 100644
index 000000000..e4b7b403c
--- /dev/null
+++ b/changelog.d/13632.bugfix
@@ -0,0 +1 @@
+Fix the running of MSC1763 retention purge_jobs in deployments with background jobs running on a worker by forcing them back onto the main worker. Contributed by Brad @ Beeper.
diff --git a/synapse/handlers/pagination.py b/synapse/handlers/pagination.py
index 74e944bce..a0c39778a 100644
--- a/synapse/handlers/pagination.py
+++ b/synapse/handlers/pagination.py
@@ -159,11 +159,9 @@ class PaginationHandler:
         self._retention_allowed_lifetime_max = (
             hs.config.retention.retention_allowed_lifetime_max
         )
+        self._is_master = hs.config.worker.worker_app is None
 
-        if (
-            hs.config.worker.run_background_tasks
-            and hs.config.retention.retention_enabled
-        ):
+        if hs.config.retention.retention_enabled and self._is_master:
             # Run the purge jobs described in the configuration file.
             for job in hs.config.retention.retention_purge_jobs:
                 logger.info("Setting up purge job with config: %s", job)