Survival of ovarian carcinoma patients undergoing second-look laparotomy after primary surgery and adjunctive chemotherapy was evaluated by retrospective chart review. From August 1976 to August 1987, 102 patients with stage I-IV disease underwent second-look laparotomy. Optimal tumor debulking and early (stage I or II) disease were positively correlated with a negative second-look laparotomy. Of the 49 patients with a "negative" second look, 15 demonstrated recurrent tumor from 12.5 to 52.5 months after laparotomy. Of the 15 recurrences, 6 were documented more than 3 years following second look. Half of the 28 patients with stage III disease and a "negative" second look have demonstrated recurrent tumor. Fifty-three patients (52%) were found to have residual disease at second-look laparotomy. Initial chemotherapy (melphalan or multiple agent) and the adequacy of primary debulking surgery (optimal vs suboptimal) were not significant factors contributing to patient survival after a positive second look. However, the size of residual disease at second-look laparotomy was a significant factor in subsequent patient survival (P less than or equal to 0.01). Fifteen patients were free of gross disease at laparotomy, but had residual tumor on microscopic examination of the specimens submitted. These patients had a 2-year actuarial survival of 78%. Forty-seven percent have survived 5 or more years after second look. Nineteen patients with tumor implants 2 cm or smaller had 2- and 5-year actuarial survivals of 61 and 31%, respectively. Nineteen patients with tumor nodules larger than 2 cm in diameter had a 2-year actuarial survival of 6%. Only 1 of 19 patients with nodules greater than 2 cm could be effectively redebulked.