All, I have a duplicate row problem and to make matters worse some tables don't have a PK or any unique identifier. See below: uid | cleanval | timestamp | received -----+-------+------------+---------- 38 | 5 | 1125081799 | 1 38 | 14 | 1122683252 | 0 38 | 5 | 1125081799 | 1 38 | 14 | 1122683252 | 0 I was hoping to have a system oid for each row but it looks like that's not the case. Anyone have any thoughts on how to remove dups? I have about 40 tables of various sizes. Cheers, Peter ---------------------------(end of broadcast)--------------------------- TIP 4: Have you searched our list archives? http://archives.postgresql.org