>There's no simple answer to your question. You can start by reviewing following topics in the BOL:
>- Deadlocking
>- Minimizing Deadlocks
>- Analyzing Deadlocks with SQL Server Profiler

I have seen that this could be cause by a non optimized index, or the absence of it on a related field. We are checking that right now. I am taking a look at the statistics reported at the table level for those _WA to see if there is something in there that we could do. I have also seen that this could be cause or be more frequent under a specific service pack of SQL Server. Also, I will have to implement a retry processus at the data class level to retry a few times on interval, such as 0.25 second, if such a situation occurs. We had that to workaround the VFPOleDB bug and that has been proven to be successful to us so far.