>Well, yeah I know, but it still seems that this could and should trigger an error . . .
>
>>I think you answered your own question. Arrays in VFP are actually one dimensional...it does some tricks to make them seem two dimensions.
>>

I'd sure call it a bug, or at the very least a documentation deficiency.
But only the latter if VFP in fact uses only the second argument to calculate the position IN A MIS-ADDRESSED array.
At the least I would have expected xxx(2,1) to display the second element.

cheers