In 1982 a randomized trial was started to compare a cisplatin-containing polychemotherapy (CAP: cyclophosphamide - CPA 750 mg/m2, adriamycin - ADM 50 mg/m2, cisplatin - P 50 mg/m2 on day 1 every 21 days) with full-dose cisplatin as single agent (P 60 mg/m2/day on days 1 and 2 every 28 days) in 44 patients undergoing exploratory laparotomy or debulking surgery for stage III-IV epithelial ovarian carcinoma with residual disease greater than 5 cm. The response was evaluated at second-look surgery with random biopsies and peritoneal washing. On the basis of the final results the authors underline some data which, although merely indicative (because of the small number of patients) appear to be worth considering since they are in accordance with the latest reports: a) similar response rate (CR + PR = 47%) to first-line treatment in the two groups; b) the CAP treatment may achieve a longer median duration of CRs than the P treatment (20 versus 11 months); c) overall survival seems similar in the two groups of patients (19 versus 18 months), whereas the survival of CRs seems longer in the CAP treated patients (greater than 32 versus 25 months). The authors also discuss some observations on a possible salvage therapy.