Piroxantrone (PRX, NSC 349174) is one of the first of a new class of intercalating agents, the anthrapyrazoles, to undergo clinical evaluation. Additionally, it is the first drug trial to prospectively test a new pharmacology-based dose escalation schema proposed for Phase I trials of anticancer compounds. In this Phase I trial, PRX was administered as a 1-h infusion every 3 weeks to patients with advanced cancer. Forty-four evaluable patients received 116 courses at doses ranging from 7.5 to 190 mg/m². The dose-limiting toxicity was myelosuppression with leukopenia predominating. Nonhematological toxicities were minimal and consisted of nausea and vomiting, alopecia, mucositis, and phlebitis. Based on this trial, the maximum tolerated and recommended Phase II doses for PRX administered on this schedule are 190 and 150 mg/m², respectively. PRX plasma elimination was rapid and best fit by a two-compartment model for 17 of 24 patients receiving ≥ 90 mg/m². The plasma clearance rate was 1290 ± 484 ml/min (720 ± 210 ml/min/m²) and did not vary with dose. The t_½α was 2.9 ± 5.3 (SD) min and the t_½β was 18.7 ± 36.5 min. Area under the concentration versus time curve (AUC) at the maximal tolerated dose (MTD) was 435 µmol·min/liter, 40% higher than the predicted AUC from preclinical testing. The percentage decrease in WBC and neutrophil count was correlated with the AUC. The potential advantage of pharmacology-based dose escalation was limited in this study by assay insensitivity, extremely rapid plasma elimination, and the proximity of the starting dose to the dose where the target AUC was achieved and standard dose escalations were to begin. Consequently, there was no reduction in the number of dose escalations required to define the maximum tolerated dose. However, the practical aspects of this approach have been established and its use is recommended for further trials where detailed preclinical pharmacological studies are available.