Thanks for responding.
Great point! However, wouldn't (01,02,03,10,12,20) sort as (01,02,03,10,12,20)? Would this concept mess it up if I had data such as (33256-0456, 33256)!!!?
Here's what I did:
1. I dropped the indexes, removed the PK and indenty, then rebuilt the table as follows:
Set Zipcode to PK, Set ZipCodeID to Identity and it works fine.
2. I wonder, with such a simple table design (and static data), if I even need the ID field. I was not planning on using the ID in a join or FK anyway. Just using zipcode in a combo box.
3. I also wonder if I there was an extra index in the original situation as when I compared the old code to the new, I saw the PK was named PK_globZipcodes2 previously!!!