>>Thanks for all the input. I tested it and sure enough it was the CP settings. I never thought of the CP as being the setting. Thanks a bunch.
>
>It's a VFP9 thing. VFP8 it was still using any old codepage, and probably lost some speed while dealing with multiple codepages in the same SQL run. VFP9 is more strict with these things.

In case when CPDBF() <> CPCURRENT() VFP8 (and VFP9 with SET ENGINEBEHAVIOR < 90) may produce incorrect result because the indexes may not match the table data after data is converted to the current code page.