OBJECTIVE Greater experience with urological laparoscopy has lead to increasing interest in its use for reconstructive surgery, such as pyeloplasty. METHODS A total of 124 cases of laparoscopic pyeloplasty were performed, of which 11 followed failed primary treatment done elsewhere, namely balloon dilation (3), endopyelotomy (3); open pyeloplasty (3), endopyelotomy plus balloon dilation (1) and open pyeloplasty plus balloon dilation (1). Nine patients had renal calculi. A 4 port, balloon dissecting, extraperitoneal laparoscopic approach was used in all except 1 patient, who had a horseshoe kidney, necessitating a transperitoneal approach. RESULTS Operative time was 29 minutes longer in the secondary pyeloplasty group compared to primary cases (173.3 vs 144.0 minutes) but the conversion rate (0% vs 1.6%) and duration of postoperative hospitalization (2.8 nights each) were no greater. The complication rate was 3.6% and 9.1%, respectively. The success rate was 98.2% and 90.9% (p = 0.63) at a mean followup of 20.2 and 19.7 months, respectively. In the 9 patients with renal calculi a total of 18 calculi (94.7%) were successfully removed and the ureter was transposed medial to a crossing vessel in 50.0%. Trainee operating did not significantly prolong the procedure vs no training (162.0 vs 143.9 minutes, p = 0.06). CONCLUSIONS Extraperitoneal laparoscopic dismembered pyeloplasty is capable of addressing all causes of ureteropelvic junction obstruction with excellent functional results and low morbidity, and with an operative time similar to that of open pyeloplasty. Secondary laparoscopic pyeloplasty does not increase hospitalization, conversion or complication rates.