Purpose: In two sequential phase ii studies, we evaluated the feasibility and efficacy of adding paclitaxel to a standard platinum/etoposide regimen in the first-line treatment of small-cell lung cancer. Patients and Methods: One hundred seventeen patients with small-cell cancer were treated between June 1993 and July 1996, The first 38 patients received a lower-dose regimen: paclitaxel 135 mg/m(2) by 1-hour infusion, carboplatin at an area under the concentration-time curve (AUG) of 5.0, and etoposide 50 mg alternating with 100 mg orally on days 1 to 10. When only mild myelosuppression was observed, doses of paclitaxel and carboplatin were increased in the subsequent 79 patients (paclitaxel 200 mg/m(2) by 1-hour infusion and carboplatin at an AUC of 6.0). All patients received four courses of treatment, administered at 21-day intervals. Patients with limited-stage small-cell lung cancer also received thoracic radiation therapy (1.8 Gy/d; total dose, 45 Gy) administered concurrently with courses 3 and 4 of chemotherapy. Results: Seventy-two of 79 patients (91%) who received the higher-dose regimen had major responses. Thirty-two of 38 (84%) with extensive-stage disease responded (21% complete response rate); median survival was 10 months for this group. With limited-stage disease, the overall response rate was 98%, with 71% complete responses; the median survival time has not been reached at 16 months. Median survival in extensive-stage patients was longer in patients who received the higher-dose regimen (10 months) than in the previous group treated with lower doses (7 months; P = .008). The higher-dose regimen was well tolerated, with myelosuppression being the major toxicity. Compared with the lower-dose regimen, grade 3/4 neutropenia increased from 8% to 38% of courses, bur the incidence of hospitalization for neutropenia and fever did not increase. Other nonhematologic toxicities were uncommon, and did not increase substantially with the higher-dose regimen. Conclusion: Paclitaxel can be added at full dose (200 mg/m(2)) to a carboplatin/etoposide combination while maintaining a tolerable toxicity profile. Median survival times in both extensive-and limited-stage patients compare favorably with other reported regimens. This regimen merits further investigation, and a randomized trial to compare this regimen with a standard carboplatin/ etoposide combination is underway. (C) 1997 by American Society of Clinical Oncology.