> 1. Jim, I DID use the NOFILTER clause, and had the same problem. Ok, so at least this is sorted out. > 2. Vlad, used( dbf() ) gave FALSE, used( cursorname ) gave TRUE. The > cursor was there when I used Disp Stat, but it couldn't be seen with file() > or in the Explorer. No wonder - Used() works with aliases only. > 3. Tom, this was a file I created with SELECT... to cursor..., not a VFP > internal file (really don't care if those are kept in memory). If you also > saw this, at least I'm not going crazy. The crucial test is if you can Append from it. If you can, then why worry. Then it means VFP decided you got tons of RAM adn didn't want to go digging down your disk. The filename is reserved in case disk would have been used (is this the proper tense? I've forgotten some of the grammar). If you can Append from, then never mind, VFP is emulating the existence of the phantom file for its own use, and yes, it had never existed.