Previous studies examining sun exposure and ocular melanoma have produced inconsistent results. We investigated this association in a population-based case-control study in Australia. Cases (n = 290) aged 18-79 years were diagnosed between January 1996 and July 1998. Controls (n = 893) were randomly selected from the electoral rolls and frequency-matched to cases by age, sex and state. A self-administered questionnaire and a telephone interview measured sun exposure on weekdays and weekends at 10, 20, 30 and 40 years of age and over the whole of life for specific jobs and recreations. Multivariate logistic regression models of ocular melanoma and sun exposure contained age, sex, region of birth, eye color and measures of ocular and cutaneous sun sensitivity as covariates. Choroid and ciliary body melanoma (n = 246) was positively associated with time outdoors on weekdays and, less persuasively, total time outdoors but not ambient solar irradiance. Odds ratios increased with increasing exposure to OR 1.8 (95% confidence interval 1.1-2.8) for the highest quarter of sun exposure on weekdays up to 40 years of age for men and women together. The strongest positive associations were for total exposure up to 40 years of age, lifetime occupational exposure and total exposure at about 20 years of age in men; all had odds ratios between 2 and 3 in the highest exposure categories. There was inconclusive evidence for an association between sun exposure and iris (n = 25) or conjunctival (n = 19) melanomas. Sun exposure is an independent risk factor for choroidal and ciliary body melanoma in Australia.