Nigel writes: > Is the PK index status UNUSABLE? could the DB crash have happened at an > inconvenient point in a batch data load, for example? Sorry, I should have included an assumption that all objects are fine. While I do have some workarounds (e.g. stats) in play, I'm more concerned with the fact that I can't use this particular solution or determine why it doesn't work. Support had me use their SQLT tool. It's much more "intrusive" than I'd normally like to use on a critical production DB, but the output it generates is impressive. Thanks! Rich -- //www.freelists.org/webpage/oracle-l