The development of new drugs in early clinical trials is currently based upon the results of preclinical antitumour and toxicity studies in animals. More recently, the use of preclinical pharmacokinetic information in mice has been proposed to also provide information that might expedite early clinical trials and more specifically phase I studies. The anthrapyrazole CI-941 was one of three chosen for phase I anticancer drug development. In addition, because of the predictability of the preclinical dose limiting toxicity and linear CI-941 pharmacokinetics in mice; a pharmacokinetically guided dose escalation scheme was attempted during the phase I trial, but had to be abandoned. 44 patients were entered who received 95 courses of treatment using a bolus injection every 21 days. The dose range was 5-55 mg/m2. The dose limiting toxicity was leucopenia and other toxicities, which included nausea and vomiting, mucositis, diarrhoea, alopecia and skin discolouration were either mild or manageable. Pharmacokinetic studies were performed with 27 courses. There were wide interpatient variations in the dose-AUC relationship (r = 0.7496) that hampered application of the proposed pharmacokinetically guided dose escalation scheme as planned. No complete or partial responses were observed. The recommended phase II dose using this schedule is 50 mg/m2.