Background:Although a causal association between mesothelioma and occupational exposure to asbestos is beyond dispute, nearly all population‐based studies of mesothelioma have found some proportion of cases with no history of asbestos exposure.Data:Incidences of mesothelioma in men and women not occupationally exposed to asbestos for the period 1980–1985 were generated from: (i) estimates of occupational exposure to asbestos in the general male population, obtained from population‐based controls in a case–control study of incident cancer in South Australia; and (ii) National Mesothelioma Surveillance Program data. The incidence in men not occupationally exposed was 8.5 per 1000000 person‐years, compared with 2.6 per 1000000 person‐years in women, a difference of 5.9 per 1000000 person‐years (90% confidence interval, 4.8–7.2).Conclusion:Mesothelioma without a history of asbestos exposure may be caused by asbestos in the general environment, but this fails to explain why such cases occur more commonly in men. Alternative explanations include the existence of another independent cause of mesothelioma, or of a co‐factor which, combined with “environmental” levels of asbestos exposure, constitutes a sufficient cause. Such a risk factor is likely to be occupational, in which case mesothelioma may be expected to occur even after occupational asbestos exposures have been eliminated.