Ovarian cancer is the leading cause of gynecological cancer-related death in Western countries. The present treatment standards for ovarian cancer are based on the association of debulking surgery with platinum-based chemotherapy. Another strategy that could be further investigated is intraperitoneal chemotherapy (IP). We previously described that the 2-h administration of intraoperative IP cisplatin did not reach satisfactory concentrations. In the present study, we present the results of a pharmacokinetic analysis performed after two consecutive 1-h IP 30 mg/l cisplatin administrations. Twenty-seven patients with advanced epithelial cancer classified FIGO stage IIIC were included in the study. Blood and IP samples were taken over a 24-h period, during and after IP treatment. Both total and ultrafiltered (Uf) platinum (Pt) concentration levels were analyzed. Biological and clinical toxicities were also recorded. With this strategy, IP Pt concentrations stayed above the target concentration (10 mg/l) for a satisfactory length of time. The serum Pt concentrations were higher than those observed with the "one-bath" protocol and they induced the occurrence of recoverable renal toxicities (3 grade 1, 7 grade 2 and 4 grade 3). The best predictive parameter for renal failure was the total Pt 24-h Area Under the Curve (AUC) with a threshold value of 25 mg h/l RR = 0.31 (95% CI 0.13 - 0.49, P < 0.01). Administration of an increased amount of cisplatin is feasible and a satisfactory level of IP Pt concentrations is obtained. However, this improvement is associated with an increase in serum Pt levels and resulting renal toxicities. An attractive solution would be to decrease Pt transfer from peritoneum to bloodstream. A phase 1 study using intraoperative IP epinephrine in order to decrease this transfer is presently being carried out.