Suramin, a drug used in the treatment of parasitic diseases, is currently being evaluated in clinical trials as an antineoplastic agent. The use of therapeutic drug monitoring and adaptive control with feedback in clinical trials of suramin was initially motivated by an association between acute neurologic toxicity and plasma suramin concentrations in excess of 350 micrograms/ml. We have prospectively examined the performance of both two- and three-compartment population pharmacokinetic models in controlling plasma suramin concentrations and have found that a three-compartment model best describes this drug. No correlation was found between the clearance of suramin and creatinine clearance, as had been previously hypothesized. The low systemic clearance of suramin and the number of parameters required to describe the three-compartment model suggest the need for a bayesian approach to the estimation of individual pharmacokinetics.