Aim of the study – Compare two medical strategies associated to surgery in women requiring for chronic pelvic pain due to stage III-IV endometriosis. Material and methods – Two different patient groups, A (N 27) and B (N 41), requiring for chronic pelvic pain, associated with AFS stage III-IV endometriosis, operated on from 1992 to 1997, were compared. The medium age was 35 and 34 years, respectively. Pelvic pain, classified in three stages, was similar in both groups but they were more AFS stage IV in group A,67 % than in group B, 46 % ( p < 0.01). Both groups had similar operative procedure: laparoscopic resection of deep endometriotic nodules or endometriomas, plus destruction of small superficial lesions using CO 2 laser (A) or bipolar coagulation (B). Associated medical strategy was different: group A: operative laparoscopy without preoperative treatment and in 25 % a second laparoscopy taking place after two-three months of LHRH analogues ; no postoperative treatment; group B, operative laparoscopy taking place after ovarian blockage with three-six weeks of Diane® (Androcur® + ethinyl estradiol), then two-three months of analogue postoperative treatment immediately followed by long term progestoid treatment in order to prevent recurrences in women without pregnancy desire. Results – After one year, 6/27 (22 %) of A and 3/ 41 (7 %) of B had no follow-up. Of the followed patients, a complete improvement was observed in 18/21 (86 %) A, 33/38 (87 %) B, moderate pelvic pain continued in 2/21 (10 %) A, 4/38 (11 %) B, and the treatment was in failure in 1/21 (5 %) A, 1/38 (3 %) B, without significant difference. After two years, 67 % of A and 76 % of B had a follow-up and the corresponding rates of complete improvement are 72 % (A), 87 % (B), incomplete improvement: 22 % (A), 10 % (B) and failure: 6 % (A), 3 % (B). The difference is lightly significant ( p < 0.05) and remains so if patients without follow-up are considered as failures. There was no persistence nor recurrence of endometriosis nor endometrioma two years after the surgery was completed. Conclusion – Since there were more stage IV endometriosis in group A than in B, the different medical strategies and particularly the long term postoperative treatment used in B seem have little influence on results. However, these data was obtained in women of medium age > 30, with a relatively short follow-up. It should be of interest to compare in a prospective multicentric study the long term follow-up of two cohorts of young women operated on for stage III-IV endometriosis, receiving or not a long term medical treatment after surgery in order to prevent recurrences.