>I'm glad I could help

Borislav

You might be interested in some tests I ran

I had previously written the code as a complicated scan through a file of 150,000 large records - your code made it much easier to add new search criteria and I was hoping also for a dramatic improvement in performance

I call my old method Scan and your code Sql

First on a 2.8 ghz with fast disks

3 selection criteria - all fields are indexed

Scan 6 seconds
Sql 2 seconds


Secondly on a networked pc - local C drive program - all fields are indexed

Scan 69 seconds
Sql 56 seconds

Rushmore reports that the Sql scans are not optimised - any comment?

Colin