>>If I select a subset of tableA using:
>>
>>select * from tableA where ...
>>
>>subsequent queries with the resulting table are 3-4 times faster than if the iniial subset is done using:
>>
>>select tableA.field1, tableA.field2 ...
>>
>>where the only difference is specifying specific fields vs. the *. I've also tried to place a nofilter on the fast query trying slow it down but it doesn't matter.
>>
>>Thanks,
>>Rock Jensen
>
>These tests are only meaningful if you shut down and restart VFP in between. when running unoptimized queries in VFP, VFP is often smart enough to create and use an index for itself (internally- ie it is not available or visible to you).


Done! I've also run into that several times so I always do my benchmarking that way.