A factory in France had used technical-grade chloromethyl-methyl-ether in the manufacture of anion exchange resins since 1958. Technical-grade chloromethyl-methyl-ether contains a human lung carcinogen, bis-chloromethyl-ether. The purpose of this cohort study was to determine if workers at the factory whose jobs had involved potential exposure to technical chloromethyl-methyl-ether had a higher incidence of lung cancer than coworkers, or others without potential exposure. Lung cancer occurred at a higher rate among potentially exposed workers than among nonexposed workers at the same plant (rate ratio (RR) = 5.0, 95% confidence interval (CI) 2.0-12.3), or among an external reference population (RR = 7.6, 95% CI 4.3-13.5). The average age at diagnosis for exposed cases was 10.5 years lower than for nonexposed cases. The predominantly small-cell cancers of the exposed were mostly oat-cell. There was a positive dose-response relation. The mean time from first exposure to diagnosis was 13 years (95% CI 8-18). Cumulative dose and induction time were not associated. The rate for the exposed workers peaked between 7 and 13 years after the start-up of the chloromethylation process, and was still above background in 1986, the end of the study period.