The association between dairy products consumption in adults and the likelihood of type 2 diabetes mellitus (T2DM) has been described, but more information on the adolescent population is needed. This nationally representative, cross-sectional school-based study aimed to describe the consumption of dairy products and their subtypes and to evaluate their association with prediabetes and T2DM in adolescents. The Study of Cardiovascular Risks in Adolescents (ERICA) includes adolescents aged 12-17 years. Dairy products consumption was evaluated by 24-h food recall. Associations with fasting glucose, glycated hemoglobin (HbA1c) and insulin resistance, as measured by homeostatic model assessment-insulin resistance (HOMA-IR), were evaluated by multivariate linear regression. Poisson regression was also used to assess the association between dairy products consumption and the combined prevalence of prediabetes and T2DM. Models were adjusted for sociodemographic, nutritional, behavioural and anthropometrics. The final sample analysed consisted of 35614 adolescents. Total intake of dairy products was inversely associated with fasting blood glucose levels after adjusting for all covariates (β = -0·452, 95 % CI -0·899, -0·005). The associations were stronger for overweight and obese adolescents. Findings were similar for full-fat dairy products and yogurt. Higher consumption of low-fat dairy products and cheese were associated with a 46 % (prevalence ratio, PR 1·46, 95 % CI 1·18, 1·80) and 33 % (PR 1·33, 95 % CI 1·14, 1·57) higher combined prevalence of prediabetes and T2DM, respectively. The total consumption of dairy products and full-fat dairy products was associated with a lower combined prevalence of prediabetes and T2DM, while the consumption of cheese and low-fat dairy products was associated with higher combined prevalence of prediabetes and T2DM in Brazilian adolescents.