>Thinking that since it was Rushmore-optimizable, I used a SCAN loop to pre-process some data for a report (was too tricky to be done with SQL) and found that there were tremendous delays in processing a relative handful of records from a fairly large table (about 100,000 records). So, I ended up SEEKing followed by a DO WHILE loop, which worked great.
>
>Is SCAN something to avoid with large tables, or did I just miss something? (My FOR expression was definitely optimizable.)

I have seen it in the past where the scan may be slow on large talbes do to the fact that the table being scan for a certain expression has an index open at the time. It could also be a memory problem as a scan seemed to go fairly quickly after reboot, but as I used my computer through the day and resources have been taken the scan may slow down. Just my guess, but I use scan end scan with a for condition and dont seem to have to many slowness problems. Ensure that the table is index on exactly what is in your for clause.