Milt, the speed of a SELECT statement is related to the size of the cursor returned and to the accurate use of indexes. If both your tables are indexed and there are only a few returned records, speed should be OK. Since you might have records from both Customer and Subscriber, I'd go with a JOIN. If your fields aren't named exactly the same (I use prefixes on fields for different tables) then use the SELECT Field1 as Myfield1, Field2 as Myfield2..... format in your SELECT statements.

HTH
Barbara

>In reference to the lookup table their will be both enteries in the subscriber table and the customer table. Their will also be multiple enteries for both - that is the invoice table could contain id for both the subscriber table and the customer table. The invoice table contains an invoice id as well as an id for the either the subscriber or the customer. The invoice id is unique for each invoice however. The lookup table needs to select the invoice number, either the subscriber id or customer id, as well as the address from either the subscriber or customer table based on the id in the invoice table - that is the customer or subscriber id. Yes you are correct that the id will never be the same for the subscriber or customer. If I use a select statement based on a large record set will this likely be slow. This is an application which needs to be fast because the users are on the phone helping clients. In that case would I be better off storing the data I need in a separate
>lookup table.