The process that is having the CXPacket problem is a SELECT statement with a hideous Where Clause. I'm not particularly sure that this is what's taking up all the CPU, as I said, but since the server is getting hammered, thought I'd try scaling back parallelism. As I said, seems to alleviated some of the CPU pressure, and the users aren't complaining quite so loudly, but I would really like to figure out what I'm actually seeing inside sp_who2 results. For instance, doesn't the 25 lines in the results set equate to how many threads SQL uses to complete the query? Or does it mean something else?

Gila: This is an OLTP system if that helps any. Found out the server have AMD 2-socket 4-core 2210 mhz processors resulting in 8 CPUs total.

Thanks to you both for helping!

Lezza