Hi This is our production stats from a 10.2 database: STATISTIC# NAME CLASS VALUE STAT_ID ---------- ---------------------------------------------- ---------- ------------------------ ---------- 332 workarea memory allocated 64 78,233 2433935387 333 workarea executions - optimal 64 266,608,529 3211650785 334 workarea executions - onepass 64 24,106 798730793 335 workarea executions - multipass 64 4,710 3804491469 NAME VALUE UNIT ---------------------------------------------- ------------------------ ------------ aggregate PGA target parameter 2,726,297,600 bytes aggregate PGA auto target 2,053,472,256 bytes global memory bound 272,629,760 bytes total PGA inuse 444,659,712 bytes total PGA allocated 900,833,280 bytes maximum PGA allocated 3,443,251,200 bytes total freeable PGA memory 153,944,064 bytes process count 141 max processes count 158 PGA memory freed back to OS 14,979,218,014,208 bytes total PGA used for auto workareas 0 bytes maximum PGA used for auto workareas 2,522,964,992 bytes total PGA used for manual workareas 0 bytes maximum PGA used for manual workareas 1,075,200 bytes over allocation count 0 bytes processed 31,447,700,993,024 bytes extra bytes read/written 7,804,559,946,752 bytes cache hit percentage 80 percent recompute count (total) 6,907,349 19 rows selected. The instance has been up for about 8 months: STARTUP_T --------- 07-AUG-09 NAME TYPE VALUE ------------------------------------ -------------------------------- ------------------------------ pga_aggregate_target big integer 2600M I was looking at the multipass executions and saw that there were 4700 of them. But then it is only 0.002% of the total work area executions. Can this statistic be safely ignored?