>So what's your opinion now? Still think IN is faster/better performance? >People always say my environment is abnormal so I wont be insulted if you >tell me this example doesnt change your opinion :D) Also, do you think >that since I'm using 2.6 tables as free tables in VFP has an effect on >these results? Someday I'll have all my tables in a VFP dbc, and thats >when I'll really put what I learned with this thread to use. TIA for your >all your help and valuable opinions. You may have you point, Rox. I guess that Exists (select ... ) works like In (select distinct...), which in turn doesn't retrieve all records, just one per key, so it should be faster. Now you've found it, explanation comes kind of natural :)