To assess the pattern of failure and survival of advanced ovarian cancer patients with microscopic residual disease at second-look following cytoreductive surgery and platinum-based chemotherapy. Nine-five women were retrospectively analyzed. Residual disease after initial surgery was > one cm in 58 (61.1%) patients, first-line chemotherapy was paclitaxel/platinum-based in 70 (73.7%) patients, second-look findings showed no macroscopic residuum but positive random peritoneal biopsies and/or positive washing ("true" microscopic residual disease) in 79 (83.2%) patients, and a macroscopic residuum which was completely resected (converted complete response) in 16(16.8%) patients. Eight-one (85.2%) patients developed recurrent disease after a median time of 14 months (range four to 51). The abdomen (29.6%) and the pelvis (28.4%) were the most common sites of failure. Two- and five-year survival after second-look were 78.1% and 31.0%, respectively. The clinical and pathological features with prognostic relevance at presentation (age, histotype, and tumor grade), as well as type of first-line chemotherapy and treatment after second-look were not related to the clinical outcome. There was a trend for a better survival in patients with optimal primary cytoreduction compared with those with suboptimal primary cytoreduction (five-year survival = 42.7% vs 23.4%). There was no significant difference in survival between the converted complete responders and the patients with "true" microscopic residual disease. These data confirm the unsatisfactory clinical outcome of patients with microscopic residual disease after first-line chemotherapy and the limited benefit of second-look reassessment.