Background: Although a causal association between mesothelioma and occ
upational exposure to asbestos is beyond dispute, nearly all populatio
n-based studies of mesothelioma have found some proportion of cases wi
th no history of asbestos exposure. Data: Incidences of mesothelioma i
n 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-b
ased controls in a case-control study of incident cancer in South Aust
ralia; and (ii) National Mesothelioma Surveillance Program data. The i
ncidence in men not occupationally exposed was 8.5 per 1 000 000 perso
n-years, compared with 2.6 per 1 000 000 person-years in women, a diff
erence of 5.9 per 1 000 000 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 expl
anations include the existence of another independent cause of mesothe
lioma, or of a co-factor which, combined with ''environmental'' levels
of asbestos exposure, constitutes a sufficient cause. Such a risk fac
tor is likely to be occupational, in which case mesothelioma may be ex
pected to occur even after occupational asbestos exposures have been e
liminated.