From 1990 to 1995, 120 consecutive patients with stage IIIC and IV ovarian carcinoma underwent surgical cytoreduction to < or = 1-cm residual disease followed by platinum-based chemotherapy. At the conclusion of chemotherapy all patients who were clinically disease free and whose CA-125 was < 35 were offered a second-look operation that obtained at least 100 tissue specimens. Of 107 patients who qualified for second look, 78 underwent the procedure. Forty-three (55.1%) had negative pathology, 20 (25.6%) were microscopically positive, and 15 (19.2%) had gross disease. Patients with positive findings received individualized salvage therapy. Patient age (P = 0.01) and the number of implants at primary surgery (P = 0.004) correlated with second-look results. Twelve (27.9%) of the patients with negative pathology have recurred. Eleven of these patients had metastatic disease > or = 10 cm at primary surgery (P = 0.003). Patients refusing second look had a median survival of 39.1 months. Approximately 60% of patients who underwent second look remain alive. Stepwise logistic regression selected two covariates significantly affecting survival: the number of implants at primary surgery (P = 0.0130) and performance of a second look (P = 0.0103). Using the protocol described in a population of optimally resected patients with advanced-stage ovarian cancer, second-look laparotomy can impact positively on survival. Patients with > 10-cm metastatic disease at primary surgery and negative second-look findings should be the focus of future protocols for consolidation chemotherapy.