A randomized trial of thymosin fraction V (60 mg/m2 s.c. twice weekly) given during induction chemotherapy and radiation therapy was performed in 91 patients with small cell carcinoma of the lung. Induction chemotherapy consisted of four cycles of an alternating combination of drugs (cyclophosphamide/Adriamycin/vincristine and cisplatin/etoposide). Radiation to the primary complex was given to patients with limited disease. All patients received prophylactic cranial irradiation. There were 35 patients with limited disease (18 randomized to thymosin and 17 to no thymosin) and 56 with extensive disease (28 thymosin and 28 no thymosin). Pretreatment immunological parameters were comparable between the two groups. For limited disease patients the overall response rate was 100%, including 66% (21 of 32) complete responders. The median duration of response was 19 mo (range, 5-57 mo) and survival 21 mo (range, 4 days to 57 mo). The 3-yr survival was 32%. For ED patients the overall response rate was 95% with 29% (13 of 48) complete. The median duration of response was 10 mo and the median duration of survival 12 mo with 13% alive at 2 yr. A comparison of the thymosin-versus no thymosin-treated patients revealed no difference in response rate, response duration, or survival whether analyzed as a whole or by extent of disease. An analysis based on pretreatment immune function and total white blood cell and absolute lymphocyte count revealed no difference in the survival distributions. No differences in the pattern of toxicity were observed between the thymosin- versus no thymosin-treated patients. The addition of thymosin fraction V during induction chemotherapy and consolidation radiotherapy did not alter outcome.