Thanks for that suggestion. I have read that before, but I don't see how it can be true in a multi-user environment. Of course, I have never tried it. I suspect it would be true if the source data was already in a table on the server, but mine is newly entered data by a user on a network client.
It seems to me, that if the external data must be located on the server, then it must be populated through the LAN which would be no different than populating the target table. If the external table is located on the client machine, it must be read through the LAN, which would also be slow. Given this environment, and since an Execute Block runs on the server as if it were a Stored Procedure, that seemed to be the best solution. After saying all of the above, I've never put either method into production, so I appreciate correcting anything I'm not understanding correctly.
