>Sergey, Thank You. That solves it!

I'd amend what SergeyB said just a little... VFP used to test that 'not a table' condition but dropped it along the way (starting at VFP5 I believe) and never really said anything when they did so.

As regards your "That solves it!", I hope you don't mean you "solved" it by setting the value to 0!

**IF** that's what you did then you didn't "solve" the problem, you just hid it from view!

There can be legitimate reasons for this condition to arise, and you may not be aware that subsequent record adds (most particularly when within TRANSACTIONs and don't know about others, but I wouldn't take a chance) will simply NOT WRITE yet WILL REPORT NO ERROR (repeatNOT WRITE yet WILL REPORT NO ERROR.
One speculative reason for the condition arising is that a write is done at a flaky sector.

In any case it is far more a case that VFP5, VFP6 and VFP7 were wrong and that VFP8 finally addressed the problem (at least to a certain extent).

good luck