i usually use only one unique id number table for all my lookup tables (every code has its own distinct number and may or may not be sequential). i think you'd need to store the codetype AND codeno the way you're numbering them now. if (for your example) positions were 1 - 3, printers 4 - 6, etc. you wouldn't need to know the codetype when looking up this info...

>Barbara,
>
>Boy have I been falling a lot lately ;)! (it's tough teaching yourself to ride that bike without training wheels or a parent to hold you til you're steady!)
>
>After reading your message I've got a question. Maybe my Allcodes table isn't set up correctly. Again the structure is:
> codetype - c12
> codeno - c5
> dispvalue - c80
>And the contents are unique by a combination of the codetype *and* the codeno. That is there are dupes in the codeno. Here (I love showing examples):
> Codetype Codeno Dispvalue
> position 00001 CIO
> position 00002 DB Manager
> position 00003 Telecom Specialist
> printer 00001 Laserjet
> printer 00002 Inkjet
> printer 00003 Dot Matrix
> software 00001 Word Perfect
> and so on......
>
>Should the codeno's all be unique so I can relate them by themselves with my data tables or is there a way to relate them with a combo-key (codetype+codeno) ... although the data tables only contain the codeno... *I* have to know which codetype it refers to and was thinking I could include the specific codetype in my filters. I'm beginning to think this isn't ideal. Your thoughts???
>
>Thanks as always,
>Mandy
>
>