OBJECTIVES: The cancer incidence among 2957 boiler welders was investigated. The subjects were registered electrical welders from 1942 to 1981. A subcohort of 606 stainless steel welders was studied separately. METHODS: The investigation was a historical prospective cohort study based on a national registry. The loss of follow up was 4.9%. RESULTS: There were 625 deaths (659 expected). There were 269 cancer cases (264 expected). An excess of lung cancer was found; 50 cases v 37.5 expected. There were three cases of pleural mesotheliomas v 1.1 expected. The subcohort of stainless steel welders had six cases of lung cancer v 5.8 expected, and one case of pleural mesothelioma v 0.2 expected. CONCLUSIONS: The welders in the study were assumed to represent a qualified work force. These welders had a small excess risk of lung cancer. The excess risk did not seem to be associated with stainless steel welding. Smoking and asbestos exposure were potential confounders.