AIMS: Basal metabolic rate (BMR) and energy imbalance are determinants of diabetes. We assessed the BMR-associated risk for diabetes in a large cohort. METHODS: Cox proportional models estimated hazard ratios (HR) for subsequent diabetes and its complications with estimated BMR among diabetes-free participants. The mediating role of vascular-metabolic biomarkers were assessed. FINDINGS: Among 341,790 participants aged 55.7 ± 8.1 years, the mean (standard deviation [SD]) estimated BMR was 1849 (234) kcal/day in men and 1340 (147) kcal/day in women. Over 12 years follow-up, 4,626 cases of diabetes accrued. Higher estimated BMR showed positive associations with diabetes, with each SD higher estimated BMR associated with a 55% increase in diabetes risk (HR 1.54, 95% confidence interval [CI] 1.49-1.59 in men; HR 1.56, 1.49-1.62 in women), and a substantive risk for cardiac complications among those with diabetes (1.71, 1.55-1.89 in men; 1.78, 1.56-2.03 in women), diabetic coma, ketoacidosis and glycaemic disturbances (1.18, 0.99-1.42 in men; 1.35, 1.21-1.50 in women), and stroke (1.27, 1.08-1.50 in men; 1.41, 1.14-1.76 in women). Glycated haemoglobin, triglycerides, high-density-cholesterol, Cystatin-C and C-reactive protein explained 10-30% of these associations. CONCLUSION: Estimated elevated BMR was strongly positively associated with risk for diabetes and its complications, suggesting to re-evaluate the role of BMR, in addition to BMI, in diabetes research.