>Hello,

>>>Yes, the same code but it won't work on my comp, I've end up with selecting
>>>to table...
>>You're SURE you have a separate cursor, not a filtered table?

>Actually, I'm not sure in anything ;). How can I check that I have separate
>cursor? If I right understand, file(dbf([tmp])) should return .f. in this
>case...

You can check with 'Disp Stat'. If you have a new cursor, it will have a name like 'c:\temp\03458761.tmp'. If it's actually a copy of your table, it will \have the same name as the table, with a different alias.

Another way is to use ? dbf('mycursorname'). Again you'll get the original table name if it's a copy.

HTH