CONCLUSIONS: To investigate the difference in the occurrence of lung cancer between foundry workers and non-foundry workers by comparing the number of workers diagnosed with lung cancer through health insurance data. METHODS: The study population was comprised of 28,884 workers who had undergone at least one general or special medical examination between January 1995 and December 1997 at the occupational health center. All of the subjects had health insurance during this period. We combined the medical examination data with the health insurance data to compare the number of foundry workers diagnosed with lung cancer and the number of non-foundry workers diagnosed with lung cancer. RESULTS: Seven workers were diagnosed with lung cancer among the 1,591 foundry workers, compared to twelve workers among the 27,293 non-foundry workers (odds ratio: 10.04, 95% confidence interval: 3.95-25.55). The seven foundry workers diagnosed with lung cancer were all exposed to dust, and six out these seven workers were engaged in finishing or shake-out processes. CONCLUSIONS: Although the information for this study was obtained from health insurance data, which has limitations such as accuracy and completeness, the number of foundry workers diagnosed with lung cancer was significantly higher than that of non-foundry workers. Therefore, a well-designed cohort study should be followed to confirm the higher lung cancer rates in foundry workers.