AIMS: Inverse associations between vitamin D status and risk of type 2 diabetes observed in epidemiological studies could be biased by confounding and reverse causality. We investigated the prospective association between vitamin D status and type 2 diabetes and the possible role of reverse causality. METHODS: We conducted a case-cohort study within the Melbourne Collaborative Cohort Study (MCCS), including a random sample of 628 participants who developed diabetes and a sex-stratified random sample of the cohort (n = 1884). Concentration of 25-hydroxyvitamin D (25(OH)D) was measured using liquid chromatography-tandem mass spectrometry in samples collected at recruitment. Logistic regression was used to estimate odds ratios (ORs) and 95% confidence intervals (CIs) for the risk of type 2 diabetes for quartiles of 25(OH)D relative to the lowest quartile and per 25 nmol/L increase in 25(OH)D, adjusting for confounding variables. RESULTS: The ORs for the highest versus lowest 25(OH)D quartile and per 25 nmol/L increase in 25(OH)D were 0.60 (95% CI: 0.44, 0.81) and 0.76 (95% CI: 0.63, 0.92; p = 0.004), respectively. In participants who reported being in good/very good/excellent health approximately four years after recruitment, ORs for the highest versus lowest 25(OH)D quartile and per 25 nmol/L increase in 25(OH)D were 0.46 (95% CI: 0.29, 0.72) and 0.71 (95% CI: 0.56, 0.89; p = 0.003), respectively. CONCLUSIONS: In this sample of middle-aged Australians, vitamin D status was inversely associated with the risk of type 2 diabetes, and this association did not appear to be explained by reverse causality.