The disposition of unchanged cisplatin was compared after two- and 24-hour intravenous (IV) infusion to eight patients with germ cell cancer (dose, 100 mg/m2), 14 patients with head and neck cancer (dose, seven patients 50 mg/m2; seven patients, 100 mg/m2). Patients were randomized to receive either a two- or 24-hour infusion in the first course of treatment and the reverse in the second course. Cisplatin renal clearance, total clearance, and the percentage of the dose excreted unchanged in urine were significantly lower with the longer infusion. Total clearance was 345 +/- 97.0 mL/min/m2 after the two-hour infusion and 268 +/- 70.7 mL/min/m2 after the 24-hour infusion (P less than .0001). Renal clearance was 79.1 +/- 35.3 mL/min/m2 and 34.1 +/- 14.9 mL/min/m2 (P less than .0001). The percentage of the dose excreted unchanged in urine was 22.9 +/- 6.5% and 12.8 +/- 4.0%, respectively (P less than .0001). The ratio of cisplatin renal clearance to creatinine clearance was 1.95 +/- .96 after the two-hour infusion and .90 +/- .40 after the 24-hour infusion (P less than .001). There was only a poor relationship between cisplatin renal clearance and creatinine clearance after a two-hour infusion (r2 = .05, P greater than .1) or 24-hour infusion (r2 = .18, P greater than .05). The severity of emesis was graded on a four-point scale and was significantly less with the 24-hour infusion than with the two-hour infusion (P less than .05). Twenty-four-hour infusion of cisplatin resulted in greater drug retention in patients due to reduced renal clearance, but was also associated with reduced emetic toxicity, probably as a result of lower peak plasma levels.