What you've been describing is n-tier design. Typically, you split the UI, business objects and data objects into separate tiers. These can be logically separated or physically separated. When you start hosting the business and data tiers on a central server (physical separation), then you can use DLLs there inside COM+. So, a DLL is used for the business and data logic.

>Thank you for the information Craig, I'll continue on with your suggestions. One question, though, since you mentioned using a DLL would be better as a best practice in this case. What type of situation might using a COM server be a best practice over a DLL?
>
>Thank you for your time,
>
>Tim