Abstract Objective Hyperglycaemia in type 2 diabetes is caused by varying degrees of two defects: low insulin sensitivity and beta-cell dysfunction. We assessed if subgrouping of patients into three pathophysiological phenotypes according to these defects could identify individuals with high or low risk of future cardiovascular events. Design This is a prospective cohort study. Methods We assessed estimates of insulin sensitivity and beta-cell function from the homeostasis model assessment-2 in 4209 individuals with recently diagnosed type 2 diabetes enrolled from general practitioners and outpatient clinics in Denmark. Individuals were followed for a composite cardiovascular endpoint (either atherosclerotic outcomes (myocardial infarction, unstable angina pectoris, stroke, coronary or peripheral revascularization), heart failure, or cardiovascular death) and all-cause mortality. Results Totally 417 individuals with the insulinopenic phenotype (high insulin sensitivity and low beta-cell function) had substantially lower risk of cardiovascular events (5-year cumulative incidence: 4.6% vs 10.1%; age-/sex-adjusted hazard ratio (aHR): 0.49; 95% CI: 0.30–0.82) compared with 2685 individuals with the classical phenotype (low insulin sensitivity and low beta-cell function), driven by atherosclerotic events. Conversely, 1107 individuals with the hyperinsulinaemic phenotype (low insulin sensitivity and high beta-cell function) had more cardiovascular events (5-year cumulative incidence: 12.6%; aHR: 1.33; 95% CI: 1.05–1.69), primarily driven by increased heart failure and cardiovascular death and increased all-cause mortality. Conclusions Simple phenotyping based on insulin sensitivity and beta-cell function predicts distinct future risks of cardiovascular events and death in patients with type 2 diabetes. These results suggest that precision medicine according to underlying type 2 pathophysiology potentially can reduce diabetes complications.