Thanks (belatedly) for the clear explanation, Steve. I've worked with SQL-Server and with native access VFP, and obviously had some misconceptions about how the ODBC worked.

Barbara

>Since VFP database is an ISAM db the ODBC driver still uses the files to access the data. When an sql select is made the odbc driver still uses the cdx if one exists to gather the data. The entire index will be sent to the client and the client machine then process the index an goes back across the network for the results. This is no different than native access. Additionally my time trials show that using odbc to access VFP data from a vfp client is slightly slower than native access. In my mind the only reason to use odbc in this scenario is to ensure database scalability. In other words, upsizing to SQL without having to change the front-end code.