The circadian rhythm-modulated delivery of anticancer drugs has been shown to reduce toxicity and improve anticancer efficacy. The aim of this phase I trial was to compare the feasibility and tolerability of carboplatin (CBDCA) administered at circadian-modulated or flat infusion rate in 24 patients with advanced cancer. Each treatment cycle consisted of a 5-day continuous intravenous infusion of CBDCA, to be repeated at 28-day intervals. Three dose levels were determined, with a CBDCA dose 15%, 40% and 60% over that calculated using Calvert's formula. Two schedules were compared: schedule A (forty-four courses), with a at circadian rhythm-modulated rate (peak at 16.00 hr) and schedule B (fifty courses), at a constant rate. At the first and second dose level neither of the administered cycles were accompanied by hematologic toxicity higher than Grade 3. At the third dose level, two cycles out of 15 for schedule A and two out of 20 for schedule B were accompanied by Grade 4 thrombocytopenia. The repeat cycles were delayed from day 28 to 42 in some patients, with no differences between circadian-modulated and flat infusion. Three partial responses out of 9 evaluable patients were observed in schedule A and 2 out of 10 evaluable patients in schedule B. We showed no potential advantage of the chronomodulated 5-day CBDCA continuous infusion method over the flat rate method. Although antitumor effects were observed in this pilot study, this treatment cannot be assessed for efficacy relative to other schedules.