Twenty-six patients with advanced renal cell carcinoma were treated with suramin administered by continuous infusion, with dosing determined by a nomogram. One patient achieved a partial response and five patients achieved a minor response or had stable disease for > 3 months. Toxicities included an immune-mediated thrombocytopenia in one patient and Staphylococcus sepsis that was not associated with neutropenia in five patients. Pharmacokinetic parameters were determined by the ADAPT II MAP-Bayesian parameter estimation program. Patient data were fit using a two-compartment open model and first-order rate elimination. This showed a wide interpatient variation in time to target level (median, 13.8 days), volume of distribution (median, 15.2 liters/m2), and t1/2-beta (median, 20.6 days). The patients who achieved a partial response, minor response, or stable disease had a slower elimination rate of suramin, compared to patients with progressive disease. Tumor specimens were obtained prior to therapy and were analyzed for the production of five different growth factor-specific RNA transcripts. These included transforming growth factor alpha, acidic fibroblast growth factor, basic fibroblast growth factor, and platelet-derived growth factor types A and B. No difference in the pattern of growth factor expression was seen in tumors of responding and nonresponding patients. Suramin does not have significant antitumor activity in renal cell carcinoma. The wide variability in pharmacokinetics suggests that individual dosing should be used in future trials of suramin for treatment for other malignancies. Pertinent corollary studies of tumor biology and clinical pharmacology should be included whenever possible in clinical trials in patients with renal cell carcinoma.