Thirteen previously untreated patients with extensive small-cell lung cancer (SCLC) were treated with the investigational agent amonafide in a dose of 300 mg/m2 intravenously (IV) over 1 hour daily for 5 consecutive days. No responses were seen in 12 eligible patients. Myelosuppression was only occasionally seen. Other toxicities included diaphoresis, chest pain, local irritation at the injection site, arthralgias, nausea and vomiting, and neuromuscular problems. There were two early deaths, both attributable to tumor progression with resultant obstruction of a vital structure. Ten patients crossed over to alternate active therapy (etoposide [VP-16]-cisplatin) and five responded. The median survival time (MST) of the whole group of treated patients was 31 weeks. In future trials of investigational new drugs in previously untreated SCLC, we recommend that patients with the following characteristics be excluded: Eastern Cooperative Oncology Group (ECOG) performance status 2, 3, and 4; superior vena cava (SVC) obstruction; any major paraneoplastic syndrome; serious comorbid illness; and extensive hepatic involvement by tumor. The trial design should include prompt crossover to active alternative therapy, such as VP-16 and cisplatin, for disease progression or for failure to respond after two treatment cycles. Also, the trial design should use an early stopping rule based on interest in identifying only very active agents with a minimum response rate of 30%.