OBJECTIVE Controversy exists concerning the timing of surgical correction of presumed ureteropelvic junction obstruction. Different opinions on the timing of intervention focus on renal function deterioration with time, poor initial relative function, or obstructive drainage curves and/or drainage time on diuretic renography. We retrospectively determined whether there is any improvement in renal function after pyeloplasty for presumed renal obstruction. METHODS We retrospectively reviewed the charts of patients who underwent pyeloplasty between 1990 and 1997 in whom preoperative and postoperative diuretic renography data were available. Patients were excluded from review when they had bilateral ureteropelvic junction obstruction, a solitary kidney, associated vesicoureteral reflux, or other bladder or ureteral abnormalities, and when preoperative and postoperative renography studies were not available. RESULTS Complete data were available for review in 79 patients 2 weeks to 18 years old (median age 6 months). Of the patients 73% were male and 73% of the affected kidneys were on the left side. Prenatal hydronephrosis had been diagnosed in 58 patients (73%), of whom 19 (33%) were observed for a variable period before pyeloplasty was performed. In all cases diuretic renography performed at the same institution using a standard protocol revealed a drainage time of 20 minutes or greater preoperatively, while in 58 cases a measurable drainage time was never achieved. As a rule, drainage improved postoperatively (mean and median 25 and 16 minutes, respectively). Open renal biopsy done at pyeloplasty in 54 patients was normal in 29. Preoperatively renal function ranged from 5 to 67% (mean and median 41 and 45, respectively). In all patients the paired t test showed no statistical difference in preoperative and postoperative renal function (p = 0.078, 95% confidence interval -3.451 to 0.185). There was no statistical change in renal function in patients with an abnormal renal biopsy regardless of the severity of renal scarring (p = 0.38) or when renal function was 40% or less (mean preoperative versus postoperative 29.7 versus 28.4%, p = 0.46). The group with greater than 40% function preoperatively had no relevant difference in function before or after surgery (mean 49.7 versus 47.8%, p = 0.065). Prenatally screened patients who were initially observed had a statistically significant difference in renal function before and after pyeloplasty (mean 45.6 versus 43%, p = 0.002). CONCLUSIONS Renal function did not improve after pyeloplasty regardless of the initial level of relative function. Renal scan revealed that differential function decreased after pyeloplasty in some patients in whom hydronephrosis was detected prenatally and who were initially followed with observation. In our opinion waiting for renal function to decrease before considering pyeloplasty is not warranted, since function does not improve even when obstruction is corrected and drainage time improves.