Jonathan, Thank you for the information. I have used the dbms_space
package before. I just wondered if there was a way to gather info for
the entire schema all at once instead of table by table.

I do suspect that we have serious performance problems and I think some
of it is due to freelist contention. I checked out the stats tables
that you mention above. Here is what I found from the v$segstat. What
does the value column quantify? milliseconds or number of waits? This
is just the first three, there were over 600 total lines.
TbspName, ObjName, STATISTIC_NAME, VALUE
PRODDTAT, F55PTRQ, buffer busy waits, 10647244
PRODDTAT, F4801, buffer busy waits, 9479677
PRODDTAT, F31122, buffer busy waits, 7851222

also in looking at the v$waitstat this is what I found
CLASS, COUNT, TIME
data block, 42546555, 65329459
undo header, 13182, 35959
undo block, 12464, 11769
extent map, 1675, 2441
segment header, 1646, 3796

definitely some segment header contention there, but quite a lot more
data block. What does that tell me other than I've got some serious
work cut out for me?
Thanks again for the help..