Hi Neil,

FYI, the SQL way is

where table3.f1 IS NULL ;

>I had to alter the code, the union all as it was giving duplicates, I also removed the DISTINCTS as per sergey's suggestion this resulted in a significant difference.
>
>Rushmore optimization was NONE on all
>
>Our code 209 recs 2.12 seconds
>Your code 209 recs 1.13 (after removing UNION ALL)
>Your code 209 recs 0.75 (after removing the DISTINCTS) not bad improvement
>
>The final code is as follows:-
>
>

>select table1.f1 from table1 left join table3 on table1.f1 = table3.f1 ;
>where ISNULL(table3.f1) ;
>union ;
>select table2.f1 from table2 left join table3 on table2.f1 = table3.f1 ;
>where ISNULL(table3.f1) ;
>order by 1 into cursor curOrphans nofilter
>

>