The purpose of this study was to gain insight into the treatment policy and survival of patients with non-small cell lung cancer (NSCLC) clinical stage IIIA in daily practice. We selected 212 patients, who had been diagnosed between 1989 and 1994 and registered by the Cancer Registry, Comprehensive Cancer Centre East (CCCE). Diagnostic tests comprised chest X-ray and bronchoscopy in all cases but one, computed tomography in 89%, mediastinoscopy in 55% and conventional tomography of the chest in 16%. NSCLC had been verified histologically in 88% and cytologically in 12%. The initial treatment for the primary tumor had been surgery alone in 13% of the patients, surgery plus radiotherapy in 8%, radiotherapy alone in 56%, chemotherapy in 1% (three patients, one in addition to surgery); 22% received none of these treatments. Median survival of the 212 patients was 9.4 months (95% confidence interval 8.3-11.0 months). Overall survival rates after 1, 2 and 3 years were 41, 17 and 8%, respectively. Three-year survival of the patients who had undergone surgery, surgery plus radiotherapy, radiotherapy alone and no treatment was 18, 19, 6 and 4%, respectively. Treatment was an independent prognostic factor (multivariate Cox's proportional hazards analysis adjusted for sub-stage, age, number of co-morbid diseases and hospital). In the same model, the Hazard rate ratio for one hospital relative to the five others was 1.9 (95% confidence interval 1.2-2.8). Surgery (whether or not in combination with radiotherapy) independently gave the best results. In conclusion, policies varied between hospitals, although the variation in overall survival was small except at one hospital. New regional management guidelines are in preparation. Physicians will be encouraged to follow these guidelines, both with regard to diagnostic tests and to treatment policies, as our study showed that differences in policy might lead to differences in survival.