Patients with inoperable non-small cell lung cancer, limited disease, were randomized to either chemotherapy (cisplatin and etoposide) or radiotherapy of the thorax. A set of questionnaires, covering 4 areas: psychosocial well-being, medical side effects, physical function, and everyday activity was filled out by 101 patients. The correlations between psychosocial well-being (dependent variable) on the one hand, and medical side effects, physical function, and everyday activity (independent variables) on the other were subjected to multiple regression analysis. No significant correlation was found between psychosocial well-being and chemotherapy- or radiotherapy-related subjective side effects. The disease-related symptoms correlated highly with psychosocial well-being during the observation period. Psychosocial well-being and the independent variables tended to be more strongly correlated during the follow-up period (14-52 weeks) than during the treatment period.