The role of fat-soluble vitamins in the pathology of type 2 diabetes needs further research. Possible protective effects could be expected for vitamins A and E via their antioxidant properties, vitamin K via its modulating effects on cytokines and insulin resistance and vitamin D via the enhancement of insulin sensitivity. However, the evidence on association between fat-soluble vitamins from diet and risk of diabetes is limited. Therefore, among 19 168 healthy Japanese of both sexes aged 40–79 years, we used the logistic regression analyses to examine the prospective association between FFQ-estimated dietary fat-soluble vitamins (A, K, E and D) and the risk of type 2 diabetes incident over a 5-year period. During this 5-year period, 494 new cases of diabetes were self-reported. Vitamins K and E from diet were associated with lowered risk of incident diabetes, whereas no associations with dietary intake of vitamin A or D were observed. The multivariable OR in the highest v. lowest quartiles of intakes were 0·71 (95 % CI 0·54, 0·93, Ptrend=0·01) for vitamin K and 0·72 (95 % CI 0·55, 0·95, Ptrend=0·02) for vitamin E. Mutual adjustment for dietary intake of these vitamins did not change the association. There were no interactions with sex, age, smoking status, BMI or having a family history of diabetes, P were >0·10. In conclusion, higher dietary intake of fat-soluble vitamins K and E, but not vitamin A or D, were associated with lowered risk of type 2 diabetes among Japanese population.