The purpose of this study was to determine the efficacy and toxicity of amonafide in unresectable or recurrent head and neck cancer and to determine if the degree of toxicity with amonafide correlated with the acetylator phenotype of the patient. Thirty patients were registered on the study and received amonafide, 300 mg/m2, over two hours each day for five consecutive days every 21 days. There was one partial response (3%) which lasted four months. The dose-limiting toxicity was myelosuppression. Acetylator phenotype was determined prior to treatment using HPLC to quantitate caffeine metabolites in urine samples after administration of caffeine. This pharmacokinetic evaluation was performed in 21 patients and revealed that (17/21) 81% of the patients were slow acetylators and 19% of the patients were rapid acetylators. No association was found between acetylator phenotype and toxicity in our patient population. Based on this study, it appears that amonafide given at 300 mg/m2 for 5 consecutive days every 21 days is not active in squamous cell carcinoma of the head and neck, and that acetylator status does not correlate with toxicity.