diff --git a/swh/scheduler/backend.py b/swh/scheduler/backend.py
--- a/swh/scheduler/backend.py
+++ b/swh/scheduler/backend.py
@@ -261,6 +261,11 @@
         select_cols = ListedOrigin.select_columns()
         insert_cols, insert_meta = ListedOrigin.insert_columns_and_metavars()
 
+        deduplicated_origins = {
+            tuple(getattr(origin, k) for k in pk_cols): origin
+            for origin in listed_origins
+        }
+
         upsert_cols = [col for col in insert_cols if col not in pk_cols]
         upsert_set = ", ".join(f"{col} = EXCLUDED.{col}" for col in upsert_cols)
 
@@ -274,7 +279,7 @@
         ret = psycopg2.extras.execute_values(
             cur=cur,
             sql=query,
-            argslist=(attr.asdict(origin) for origin in listed_origins),
+            argslist=(attr.asdict(origin) for origin in deduplicated_origins.values()),
             template=f"({', '.join(insert_meta)})",
             page_size=1000,
             fetch=True,
diff --git a/swh/scheduler/tests/test_scheduler.py b/swh/scheduler/tests/test_scheduler.py
--- a/swh/scheduler/tests/test_scheduler.py
+++ b/swh/scheduler/tests/test_scheduler.py
@@ -686,6 +686,15 @@
 
         assert all(origin.first_seen == origin.last_seen for origin in ret)
 
+    def test_record_listed_origins_with_duplicate(self, swh_scheduler, listed_origins):
+        # the duplicates must be in the same page to raise the "on conflict error"
+        listed_origins.insert(0, listed_origins[0])
+
+        ret = swh_scheduler.record_listed_origins(listed_origins)
+
+        # without the duplicate
+        assert len(ret) == len(listed_origins) - 1
+
     def test_record_listed_origins_upsert(self, swh_scheduler, listed_origins):
         # First, insert `cutoff` origins
         cutoff = 100