The pharmacokinetics of the cisplatin (DDP) species detected by measurement of diethyldithiocarbamate (DDTC)-reactive species (DDTC-DDP) were compared to the pharmacokinetics of the species detected by measurement of total ultrafilterable platinum in patients receiving DDP alone or in combination with the nephroprotective agent sodium thiosulfate. The doses of DDP studied were 100 mg/m2 (11 courses given to eight patients) and 202.5 mg/m2 (five courses given to four patients) given as 2 h i.v. infusions, the latter with concurrent thiosulfate. When DDP was given alone (100 mg/m2) the two assays yielded the same area under the curve (AUC) values for DDTC-DDP and total ultrafilterable platinum during the first 4 h after the start of infusion; however, beyond 4 h post-infusion, the AUC for total ultrafilterable platinum was consistently greater than that for DDTC-DDP. When DDP was given with thiosulfate (202.5 mg/m2), the AUC for total ultrafilterable platinum was significantly greater than that of DDTC-DDP during the whole sampling period. The ratio of the AUC for total ultrafilterable platinum to DDTC-DDP, when DDP was given with thiosulfate, was barely significantly greater than that when DDP was given alone. These data indicate that during and immediately following a short infusion of DDP the major platinum-containing species present in plasma ultrafiltrate are still capable of reacting with nucleophilic sites on molecules such as DDTC; however, as the reactive species are eliminated, longer half-lived non-reactive ultrafilterable platinum species begin to predominate. They also indicate that although thiosulfate does neutralize a measurable amount of DDP in the plasma on the schedule employed, this degree of neutralization is not sufficient to explain the protection against DDP-induced nephrotoxicity produced by thiosulfate.