Hi, I have the following data: >select min(start_air_date) start_date, max(end_air_date) end_date , count(*) num_units from br_bbdef_unit where brdbuy_def_id = 352864 group by start_air_date, end_air_date order by 1; START_DAT END_DATE NUM_UNITS --------- --------- ---------- 07-JAN-08 11-JAN-08 1 14-JAN-08 18-JAN-08 1 21-JAN-08 25-JAN-08 1 28-JAN-08 01-FEB-08 1 04-FEB-08 08-FEB-08 3 11-FEB-08 15-FEB-08 1 18-FEB-08 22-FEB-08 1 25-FEB-08 29-FEB-08 1 03-MAR-08 07-MAR-08 1 10-MAR-08 14-MAR-08 1 What I would like to do is to group consecutive dates with the same number of units together to get their min/max values. So in this case I would end up with 3 records: Start Date End Date Units/Wk 07-jan-08 01-feb-08 1 04-feb-08 08-feb-08 3 11-feb-08 14-mar-08 1 I'm thinking analytic functions may help here, but I haven't been able to figure out a query to produce the desired output. Any help would be greatly appreciated. Thanks, ----------------------------------- Alan Davey Senior Developer Advertising Systems Group Harris Corporation, Broadcast Communications 4 Century Drive Parsippany, NJ 07054 -- //www.freelists.org/webpage/oracle-l