Findings on dietary glycaemic index (GI) and glycaemic load (GL) as risk factors for type 2 diabetes have been controversial. We examined the associations of dietary GI and GL and the associations of substitution of lower-GI carbohydrates for higher-GI carbohydrates with diabetes risk in a cohort of Finnish men. The cohort consisted of 25 943 male smokers aged 50-69 years. Diet was assessed, at baseline, using a validated diet history questionnaire. During a 12-year follow-up, 1098 incident diabetes cases were identified from a national register. Cox proportional hazard modelling was used to estimate the risk of diabetes, and multivariate nutrient density models were used to examine the effects of substitution of different carbohydrates. Dietary GI and GL were not associated with diabetes risk; multivariate relative risk (RR) for highest v. lowest quintile for GI was 0·87 (95 % CI 0·71, 1·07) and for GL 0·88 (95 % CI 0·65, 1·17). Substitution of medium-GI carbohydrates for high-GI carbohydrates was inversely associated with diabetes risk (multivariate RR for highest v. lowest quintile 0·75, 95 % CI 0·59, 0·96), but substitution of low-GI carbohydrates for medium- or high-GI carbohydrates was not associated with the risk. In conclusion, dietary GI and GL were not associated with diabetes risk, and substitutions of lower-GI carbohydrates for higher-GI carbohydrates were not consistently associated with a lower diabetes risk. The associations of dietary GI and GL with diabetes risk should be interpreted by considering nutritional correlates, as foods may have different properties that affect risk.