Faster response time of queries

Question

I have a query which joins a few tables with seveeral thousand rows each. This query normaly returns tens of thousands and the response time is almost 10 minutes and it's not acceptable for web application.
To speed it up I just want Oracle to return the first let's say 1000 rows.
Changing the max rows returned parameter(APEX) to 1000 doesn't help at all. It seems like query executes in full and then only the first 1000 rows of the resultset are sent.
So my question is: is there way to instruct Oracle to stop execution of the query once first n rows are retrieved?
I tried the SELECT /* FIRST_ROWS(1000) */ .... but this doesn't help. and I wonder how could it when it seems that TOAD determines this as a comment and doesn't change the optimizer mode - still ALL_ROWS.
What am I doing wrong here, this is the first time I am trying to use FIRST_ROWS hint , - is there another - better way to speed up my query?

Answer 1

Hi Bob, thanks for the response. rownum < n was the first thing I tried. One would think that if a query takes 5 minutes to execute and returns 50 000 rows then after adding rownum < 5000 it shuldn't take more than a minute - well it takes pretty much the same time as w/out rownum < n. It seems like rownum is determined for the whole resultset and then the where codition is applied.
The tables actually have much more than a few thousand rows. one is with close to 250 000 and couple other tables with over a million and I don't see much that I can optimize. I think being able to only return certain first n rows fast for web applications must be fairly common situation when dealing with large tables/views.