diff --git a/brozzler/cli.py b/brozzler/cli.py
index 188d591..458e447 100644
--- a/brozzler/cli.py
+++ b/brozzler/cli.py
@@ -2,7 +2,7 @@
 '''
 brozzler/cli.py - brozzler command line executables
 
-Copyright (C) 2014-2017 Internet Archive
+Copyright (C) 2014-2019 Internet Archive
 
 Licensed under the Apache License, Version 2.0 (the "License");
 you may not use this file except in compliance with the License.
@@ -605,6 +605,10 @@ def brozzler_purge(argv=None):
             '--site', dest='site', metavar='SITE_ID', help=(
                 'purge crawl state from rethinkdb for a site, including all '
                 'pages'))
+    group.add_argument(
+            '--finished-before', dest='finished_before', metavar='YYYY-MM-DD',
+            help=('purge crawl state from rethinkdb for a jobs that ended '
+                  'before this date'))
     arg_parser.add_argument(
             '--force', dest='force', action='store_true', help=(
                 'purge even if job or site is still has status ACTIVE'))
@@ -653,6 +657,20 @@ def brozzler_purge(argv=None):
                         '(override with --force)', site_id)
                 sys.exit(1)
         _purge_site(rr, site_id)
+    elif args.finished_before:
+        finished_before = datetime.datetime.strptime(
+                args.finished_before, '%Y-%m-%d').replace(
+                        tzinfo=doublethink.UTC)
+        reql = rr.table('jobs').filter(
+                r.row['finished'].default(r.maxval).lt(finished_before).or_(
+                    r.row['starts_and_stops'].nth(-1)['stop'].default(r.maxval).lt(finished_before)))
+        logging.debug(
+                'retrieving jobs older than %s: %s', finished_before, reql)
+        for job in reql.run():
+            # logging.info('job %s finished=%s starts_and_stops[-1]["stop"]=%s',
+            #         job['id'], job.get('finished'),
+            #         job.get('starts_and_stops', [{'stop':None}])[-1]['stop'])
+            _purge_job(rr, job['id'])
 
 def _purge_site(rr, site_id):
     reql = rr.table('pages').between(