The objective of the study is to determine whether surgery influences the outcome of stage IV ovarian cancer. The study design is as follows: From May 1995 to December 2000, 129 patients with FIGO stage IV ovarian cancer, recruited in 42 centers, were prospectively included in GINECO first-line randomized studies of platinum-based regimens with paclitaxel administered simultaneously or sequentially. In all, 109 were eligible for this study. Standard peritoneal cytoreductive surgery was defined as a procedure including at least total hysterectomy, bilateral salpingo-oophorectomy, omentectomy, and peritoneal debulking. Surgery was considered optimal if residual lesions were smaller than 1 cm. The Kaplan-Meier method was used to compare survival. Initial abdominopelvic cytoreductive surgery was considered standard in 55 (54%) patients. Abdominopelvic surgery was optimal in 29 patients and nonoptimal in 26. Twenty-two (22%) patients had a simple biopsy, and 25 (24%) patients underwent substandard surgery. Twenty-two of these 47 patients without initial standard surgery underwent a second surgical procedure, and 17 of the 22 patients completed standard surgery. The median overall survival time in the entire population was 24.3 months (95% confidence interval [CI], 19.5-29.1 months). Patients treated without a cytoreductive surgical procedure had significantly worse median survival (15.1 months; 95% CI, 5.4-24.9 months) than patients who had optimal primary surgery (22.9 months; 95% CI, 15.6-30.1 months), nonoptimal primary surgery (27.1 months; 95% CI, 21.2-32.9 months), or neoadjuvant chemotherapy followed by surgery (45.5 months; 95% CI, 23.5-67.5 months) (P= .001). In conclusion, this study shows a significant benefit of debulking surgery in stage IV ovarian cancer patients who responded to neoadjuvant chemotherapy. Neoadjuvant chemotherapy can help to select patients for surgery.