Ovarian carcinoma demonstrates a steep dose-response curve for cisplatin, but even very small levels of acquired resistance at the cellular level are sufficient to block the efficacy of intravenous (IV) cisplatin. The intraperitoneal (IP) route of administration produces a 12-fold to 15-fold greater exposure for the peritoneal cavity, and concurrent use of IV thiosulfate permits the safe IP injection of 200 mg/m2 cisplatin. In this study, two phase II trials of an IP regimen containing cisplatin 200 mg/m2 and etoposide 350 mg/m2 with IV thiosulfate were conducted; the first trial enrolled patients with residual disease less than 2 cm who had failed primary cisplatin-based IV chemotherapy and the second trial newly diagnosed ovarian carcinoma patients irrespective of the size of residual disease after primary surgery. As salvage therapy, the IP cisplatin/etoposide regimen produced a median survival of 26 months from the start of IP therapy and 51 months from diagnosis. As first-line therapy, the median survival has not yet been reached; projected survival is 68% at 27 months. In both studies the major toxicity was myelosuppression; the use of concurrent thiosulfate almost completely eliminated serious nephrotoxicity and neurotoxicity. The size of the largest tumor mass was an important determinant of efficacy in both settings. The results of these trials are consistent with the hypothesis that increased drug delivery will result in higher response rates and improved survival. Data are sufficiently encouraging to mandate phase III randomized trials of this program.