>>Sergey,
>>
>>>It's actualy opposite. IN subquery runs once and than result is used by the main query. The EXISTS uses correlated subquery so it has to run for every record in the main query.
>>
>>This test code would seem to say that IN is doing a lot more work than EXISTS.
>>
>That's because you put f2() function in the wrong place in the EXISTS subquery. See revised code below.

CLEAR
>create cursor x1 ( i1 i, c1 c(10) )
>for i = 1 to 10000
>   insert into x1 values ( i, transform( i ) )
>endfor
>index on i1 tag i1
>
>create cursor x2 ( i1 i, c1 c(10) )
>for i = 1 to 10000 step 2
>   insert into x2 values ( i, transform( i ) )
>endfor
>index on i1 tag i1
>
>n1 = 0
>n2 = 0
>
>select *, f1() ;
>   from x1 where exists ( select i1  from x2 WHERE f2() AND x2.i1 = x1.i1 ) ;
>   into cursor x3
>
>? n1, n2
>
>n1 = 0
>n2 = 0
>
>select *, f1() ;
>   from x1 where i1 in ( select f3(i1) from x2 ) ;
>   into cursor x3
>
>? n1, n2
>
>function f1()
>n1 = n1 + 1
>return
>
>function f2()
>n2 = n2 + 1
>RETURN
>
>function f3(tnVal)
>n2 = n2 + 1
>RETURN tnVal
>

However, this test only shows how many records where processed by query and subquery. I doesn't take in account other factors. If you run speed tests you'll find out that in VFP IN subquery is faster in most cases than EXISTS.

Sergey,
Didn't do the tests. You might be right. Just thinking on SQL specification most of the time EXISTS sounds to be faster to me. Both subqueries might be called N times, however say subquery for IN is returning 1000 rows and the value compared is about the ends. Each time nearly 1000 rows should be checked. OTOH for EXISTS it's either a .T./.F. return. Subquery might still return 1000 rows but 1 or 1000 means .T. and returns immediately.
Just my 0.02 cents.
Cetin