Studies of the relationship between circulating 25-hydroxyvitamin D (25(OH)D) and cancer risk have been inconsistent. We hypothesized that serum 25(OH)D was associated with total non-skin cancer incidence and mortality, and/or specifically with colorectal, lung, breast or prostate cancer in an Australian cohort. Serum 25(OH)D was measured in 3818 participants (2166 females) in the 1994/1995 Busselton Health Survey aged 25 to 84 years at baseline. Cancer mortality and events over 20 years follow-up were determined by data linkage. The mean serum 25(OH)D concentration was 60.6 ± 18.0 nmol/L, with 28%, 54% and 18% of participants in the lower (<50 nmol/L), middle (50-75 nmol/L) and higher (≥75 nmol/L) vitamin D status groups, respectively. During follow-up (excluding the first 2 years), 212 participants died from non-skin cancer and 634, 110 and 44 participants had non-skin, colorectal and lung cancer events, respectively; 113 women had breast cancer and 122 men had prostate cancer events. For colorectal cancer, lower circulating 25(OH)D was associated with significantly higher risk compared with the middle group (covariate-adjusted HR 1.62, 95% CI 1.04, 2.53). For breast cancer, women with a higher 25(OH)D level had lower risk than women in the middle group (HR 0.38, 95% CI 0.16, 0.89) and the lower group (HR 0.37, 95% CI 0.15, 0.89). Serum 25(OH)D was not associated with overall cancer death or event, or with lung or prostate cancer. In this community-based cohort, lower 25(OH)D levels were associated with increased risk of colorectal and breast cancer, but not overall cancer risk.