The pyrazoloacridine (PZA) analogue NSC366140 (PD115934) entered clinical trial based on unique preclinical characteristics including solid tumor selectivity in vitro, marked antitumor activity in vivo against murine solid tumors, selectivity against noncycling cells, and activity against multidrug-resistant tumor cells. After identification of the pre-clinical efficacy and an acceptable toxicity profile, a Phase I study of PZA was carried out. A total of 28 patients was entered and received a total of 67 treatment courses. The drug was administered via a 1-h infusion every 21 days. The starting dose was 30 mg/m2 with 2-fold dose escalations through 480 mg/m2. The next dose escalation was 50%, to 720 mg/m2. Grade I through grade IV toxicities were observed. Since no dose-limiting toxicities were observed at 480 mg/m2, and up to grade IV toxicities were observed at 720 mg/m2, an intermediate dose, 600 mg/m2, was evaluated. Dose-limiting toxicities at 720 mg/m2 were hematological (grade III and IV neutropenia) in four of six patients and neurological (up to grade III cerebral toxicities, including restlessness, dizziness, agitation/anxiety, personality changes, and nightmares, as well as myoclonus) in three of six patients treated. The pharmacokinetic parameters which helped predict these toxicities included area under the curve and peak plasma level. Pharmacokinetic studies showed interpatient variations in all parameters studied. The mean area under the curve levels of PZA at the highest two dose levels in patients were near the level detected in mice at their maximum tolerated total dose. The recommended starting dose for Phase II trials using this schedule is 600 mg/m2.