Hi Dave, DD> It looks like you are getting a sequential scan instead of an index DD> scan. What is your locale setting? As far as I know Postgres doesn't DD> support using indexes with LIKE unless you are using the C locale. Actually no, I am using de_DE as locale because I need the german order-by support. But even for a seq-scan it seems pretty slow, but that's just a feeling. The table currently has ~172.000 rows and is suposed to rise to about 1 mio or more. Is there any way to speed the like's up with a different locale than C or to get an order by in a different Locale although using the default C locale? DD> Also, in the future you only need to post EXPLAIN ANALYZE not EXPLAIN DD> ANALYZE VERBOSE. ok, i will keep that in mind :-) didn't know how verbose you would need it *smile* Best regards Manuel