The same procedure, minus the refcursor but called with the other parameters exactly the same, can fetch the data in less than a second. A 10046 trace on both methods shows the procedure goes through a lot more work to get the same 70 rows with the refcursor. The procedure (which I did not write) has a bunch of cardinality hints that were apparently required to make it perform worth diddly. Is there something about refcursors that can cause such hints to have no effect? ---------------------------------------------------------------- Please see the official ORACLE-L FAQ: http://www.orafaq.com ---------------------------------------------------------------- To unsubscribe send email to: oracle-l-request@xxxxxxxxxxxxx put 'unsubscribe' in the subject line. -- Archives are at //www.freelists.org/archives/oracle-l/ FAQ is at //www.freelists.org/help/fom-serve/cache/1.html -----------------------------------------------------------------