You're right that reindexing might be the best - the reason we hacked the indexes then instead of reindexing was because we were working with relatively slow computers (Hey, this was 5+ years ago - a 386 was fast) and huge tables.
>I considered that, but never looked at the CDX to see how easy it would be. I dunno, seems to me the simplest thing to do is create a new datasession, open up the table with the assumed alias (the DBF name I'm guessing) and INDEX on from there.