OBJECTIVE A Japanese randomized trial showed superior survival for patients with extensive-disease (ED) small-cell lung cancer (SCLC) receiving irinotecan plus cisplatin compared with etoposide plus cisplatin. The present trial evaluated the efficacy of irinotecan plus carboplatin (IC) compared with oral etoposide plus carboplatin (EC). METHODS Patients with ED SCLC were randomly assigned to receive either IC, which consisted of carboplatin (area under the curve = 4; Chatelut formula) and irinotecan (175 mg/m2) intravenously both on day 1, or EC, which consisted of carboplatin as in IC and etoposide (120 mg/m(2)/d) orally on days 1 through 5. Courses were repeated every 3 weeks with four cycles planned. Doses were reduced by one third in patients with a WHO performance status (PS) of 3 to 4 and/or age older than 70 years. Primary end point was overall survival (OS). Secondary end points were quality of life (QOL) and complete response (CR) rate. RESULTS Of 220 randomly assigned patients, 209 were eligible for analysis (IC, n = 105; EC, n = 104). Thirty-five percent were older than 70 years, and 47% had a PS of 2 to 4. The groups were well balanced with respect to prognostic factors. OS was inferior in the EC group (hazard ratio = 1.41; 95% CI, 1.06 to 1.87; P = .02). Median survival time was 8.5 months for IC compared with 7.1 months for EC. One-year survival rate was 34% for IC and 24% for EC. CR was seen in 18 IC patients compared with seven EC patients (P = .02). There were no statistically significant differences in hematologic grade 3 or 4 toxicity. Grade 3 or 4 diarrhea was more common in the IC group. QOL differences were small, with a trend toward prolonged palliation with the IC regimen. CONCLUSIONS IC prolongs survival in ED SCLC with slightly better scores for QOL.