Plasma branched-chain amino acids (BCAAs) are linked to metabolic disease, but their relevance for prediction of type 2 diabetes development is unclear. We determined the association of plasma BCAAs with type 2 diabetes risk in the prevention of renal and vascular end-stage disease (PREVEND) cohort. The BCAAs were measured by means of nuclear magnetic resonance spectroscopy. We evaluated the prospective associations of BCAAs with type 2 diabetes in 6244 subjects. The BCAAs were positively associated with HOMA-IR after multivariable adjustment (p < 0.0001). During median follow-up for 7.5 years, 301 cases of type 2 diabetes were ascertained. The Kaplan-Meier plot demonstrated that patients in the highest BCAA quartile presented a higher risk (p log-rank < 0.001). Cox regression analyses revealed a positive association between BCAA and type 2 diabetes; the hazard ratio (HR) for the highest quartile was 6.15 (95% CI: 4.08, 9.24, p < 0.0001). After adjustment for multiple clinical and laboratory variables, the association remained (HR 2.80 (95% CI: 1.72, 4.53), p < 0.0001). C-statistics, Net reclassification improvement, and −2 log likelihood were better after adding BCAAs to the traditional risk model (p = 0.01 to <0.001). In conclusions, high concentrations of BCAAs associate with insulin resistance and with increased risk of type 2 diabetes. This association is independent of multiple risk factors, HOMA-IR and β cell function.