In a single-institution retrospective cohort study, 230 patients were treated for stage III primary ovarian cancer and 175 became eligible for second-look operations by virtue of a complete clinical response after primary surgical cytoreduction and platinum-based combination chemotherapy. Of these, 109 underwent a second-look operation. Optimal primary cytoreduction was defined as residual disease < or =1 cm. Median follow-up was 68.3 months. Five-year survival for all the 230 stage III ovarian cancers was 43.4%. Among all eligible patients (n = 175), there was no survival difference (P = 0.67) in those having second look (57.3%, 5-year survival) versus no second look (48.7%). In those patients with optimal primary cytoreduction (n = 118), there was no survival advantage to second look (69% versus 61%, P = 0.7). However, in those with suboptimal primary cytoreduction (n = 47), 5-year survival was 36% in those having second look versus only 13% in those refusing second look (P < 0.05). Multivariate analysis identified second-look surgery as the only significant independent prognostic variable affecting survival (RR = 0.321, P < 0.04). Patients with suboptimal debulking at primary surgery for stage III ovarian cancer appear to achieve a survival benefit from second-look surgical procedures, presumably from the early identification and treatment of residual disease.