diff --git a/podcasts/discover.py b/podcasts/discover.py
index e526f23..bc12489 100644
--- a/podcasts/discover.py
+++ b/podcasts/discover.py
@@ -281,10 +281,11 @@ def topics_of(user, days=7):
     [{'label', 'ids'}], the largest first."""
     from feeds import textproc
     since = timezone.now() - timedelta(days=days)
+    # Titles only: the shownotes bring every podcast's furniture ("Link", "Abo", "Mediathek").
     eps = list(PodcastEpisode.objects.filter(feed__user=user, feed__kind=PodcastFeed.PODCAST, pub_date__gte=since)
                .exclude(feed__new_episodes=PodcastFeed.NEW_IGNORE)
-               .values_list('id', 'title', 'description', 'feed__title'))
-    stories = [(i, f'{t}. {_plain(d)[:400]}', {ft}) for i, t, d, ft in eps]
+               .values_list('id', 'title', 'feed__title'))
+    stories = [(i, t, {ft}) for i, t, ft in eps]
     return textproc.topics(stories, max_topics=10)
 
 
diff --git a/podcasts/test_lists.py b/podcasts/test_lists.py
index 6fe4c19..86031d0 100644
--- a/podcasts/test_lists.py
+++ b/podcasts/test_lists.py
@@ -81,6 +81,14 @@ class MerklisteTests(Base):
         self.post('/podcasts/progress/save/', {'episode_id': self.eps[2].id, 'position_seconds': 100})
         self.assertEqual(self.later(), {self.eps[2].id}, 'started stays, heard goes')
 
+    def test_audiobooks_go_on_neither_list(self):
+        book = PodcastFeed.objects.create(user=self.user, rss_url='https://example.org/book', title='Buch',
+                                          kind=PodcastFeed.AUDIOBOOK)
+        ch = PodcastEpisode.objects.create(feed=book, guid='c1', title='Kapitel 1', audio_url='https://example.org/c1.mp3')
+        self.post('/podcasts/queue/add/', {'episode_ids': [ch.id, self.eps[0].id]})
+        self.post('/podcasts/later/add/', {'episode_ids': [ch.id]})
+        self.assertEqual((self.queue(), self.later()), ([self.eps[0].id], set()))
+
     def test_discard_leaves_both_lists(self):
         self.post('/podcasts/later/add/', {'episode_ids': [self.eps[0].id]})
         self.post('/podcasts/queue/add/', {'episode_ids': [self.eps[0].id]})
diff --git a/podcasts/views.py b/podcasts/views.py
index 3d2a9e7..f4fd573 100644
--- a/podcasts/views.py
+++ b/podcasts/views.py
@@ -95,10 +95,19 @@ def _current_episode_id(user, queued, hint=None):
     return p
 
 
+def _not_books(episode_ids):
+    """The ids that aren't an audiobook's chapters: those are played in the library, in their
+    book's order, and never sit in the Wiedergabeliste or the Merkliste (docs/audiobooks.md)."""
+    books = set(PodcastEpisode.objects.filter(id__in=episode_ids, feed__kind=PodcastFeed.AUDIOBOOK)
+                .values_list('id', flat=True))
+    return [i for i in episode_ids if i not in books]
+
+
 def _enqueue(user, episode_ids, where, current_hint=None):
     """Put episodes that aren't queued yet into the queue at `where`
     (back, front or after_current), keeping their given order.
     Returns the ids that were added."""
+    episode_ids = _not_books(episode_ids)
     queued = _queue_order(user)
     present = set(queued)
     new = []
@@ -199,7 +208,7 @@ def _later_add(user, episode_ids, auto=False):
     """Put episodes on the user's Merkliste (those not on it yet). Like queueing they are no
     longer new; wanted again after "Folge löschen". Returns the ids added."""
     have = set(PodcastLater.objects.filter(user=user, episode_id__in=episode_ids).values_list('episode_id', flat=True))
-    new = [i for i in dict.fromkeys(episode_ids) if i not in have]
+    new = [i for i in dict.fromkeys(_not_books(episode_ids)) if i not in have]
     if not new:
         return []
     PodcastLater.objects.bulk_create([PodcastLater(user=user, episode_id=i, auto=auto) for i in new], ignore_conflicts=True)