>>

SELECT state FROM table2 ;
>>  WHERE State NOT IN (SELECT state FROM table1)
>>* or using correlated query
>>SELECT state FROM table2 ;
>>  WHERE NOT EXISTS (SELECT state FROM table1 WHERE state = table2.state )
>>
>>

>
>While VFP might optimize a correlated subquery like yours, in generally, when you can write a readable query without a correlated subquery, it's a better choice. (To the best of my knowledge, any correlated EXISTS query can be rewritten as an IN query, but sometimes the result is pretty obscure.)
>
>Tamar

I've to disagree with you here because wich query is faster depends on the query itself and data it is ran against. I don't know about any tests that shows that query without a correlated subquery is always faster. Also if you move to Sql Server you will find that correlated subquery is recomended over other approch because it's faster in most cases.