>>Here's one more for you
>>

>>select * FROM PARENT p
>>	JOIN (SELECT distinct PK_FLD FROM CHILD WHERE(&cChildFilter)) dt1
>>		ON dt1.PK_fld = p.PK_fld
>>

>>
>
>I think I see where your approach is better. In your SQL Select, VFP has to "query" CHILD table only once. In my approach, SQL Select has to "jump" from quering PARENT to CHILD and back.

Hi Dmitry,

In VFP the queries with JOIN usualy faster than with correlated EXISTS subqueries. The opopsite is true for MS SQL Server.