>>Nah. REINDEX only reindexes using the information in the table header. That is why on many other threads here, we all advise to DELETE TAG ALL first then create your indexes to eliminate bloat and corruption if there was any.
>
>Yes, I've seen these before, so I never use Reindex...but this is a different situation, where I just do consecutive (with some processing in between) programmatic Index assignments without any table structure modifications. This occurred several months ago, also, but I didn't try the Delete Tag idea the first time, I just added a different extra tag name then.
>
>After the 2nd Index (no Delete Tag), what occurs is that sometimes vfp can't "see" duplicate keyed records. I.e., I have 8 records in the browse window, but only two are unique by index. A Scan runs with only two passes, one for each unique key. With the Delete Tag between Indexes, all 8 are processed as should be...
>
>Well, I know the fix, so perhaps I shouldn't dwell on this problem...a little interesting, though...

But, your experience is nice to know for future reference because I have done what you have before without realizing I may not be getting what I expect.