Clay Colburn wrote: > For those not inclined to run through the example, the answer was that it > does not update the row, oracle intelligently skips the operation. Can you post your example? I've tried several tests where the data is same/different, before/after commit, and multiple columns -- all unindexed -- but am unable to find a correlation in the numbers resulting from Jared's query run before and after each test. It's probably something obvious that I'm not seeing. It *is* Monday after all... Rich -- //www.freelists.org/webpage/oracle-l