>>Hi Ray,
>>
>>sorry, took me a while to reply.
>
>That's OK Christof, I've been occupied all week with training customer's half way across the country and just returned last night.
>
>Where are you suggesting the call to the Sleep function should go? Somewhere inside my MyTableUpdate method or more globally?

Oh, nevermind <g>; I think I see where you're going with this: call Sleep just prior to the TABLEUPDATE to make sure the two machines don't hit TABLEUPDATE at the same time? But if you just move the call around randomly, what's to prevent the two users from just happening to hit the keyboard at times separated by the same time interval generated by the RAND() and still calling TABLEUPDATE at the same time?

The open-a-second-instance-in-another-work-area idea might be worth a shot; I'll give that a try if my customer is still having issues after trying different delay and max try values.