AIMS: Whether glycaemic control is associated with cardiovascular outcomes in patients with type 2 diabetes (T2D) is unclear. Consequently, we assessed the relationship between glycated haemoglobin (HbA1c ) and cardiovascular outcomes in a placebo-controlled randomized trial which demonstrated no cardiovascular effect of sitagliptin in patients with T2D and atherosclerotic vascular disease. METHODS AND RESULTS: Secondary analysis of 14 656 TECOS participants with time to event analyses using multivariable Cox proportional hazard models. During a median 3.0 (interquartile range 2.3-3.8) year follow-up, 456 (3.1% of 14 656) patients had first hospitalization for heart failure (HF), 1084 (11.5%) died, 1406 (9.6%) died or were hospitalized for HF, and 1689 (11.5%) had a non-HF cardiovascular event (cardiovascular death, non-fatal stroke, non-fatal myocardial infarction, or hospitalization for unstable angina). Associations between baseline or time-varying HbA1c and cardiovascular outcomes were U-shaped, with the lowest risk when HbA1c was around 7%. Each one-unit increase in the time-varying HbA1c above 7% was associated with an adjusted hazard ratio (HR) of 1.21 [95% confidence interval (CI) 1.11-1.33] for first HF hospitalization, 1.11 (1.03-1.21) for all-cause death, 1.18 (1.09-1.26) for death or HF hospitalization, and 1.10 (1.02-1.17) for non-HF cardiovascular events. Each one-unit decrease in the time-varying HbA1c below 7% was associated with an adjusted HR of 1.35 (95% CI 1.12-1.64) for first HF hospitalization, 1.37 (1.16-1.61) for death, 1.42 (1.23-1.64) for death or HF hospitalization, and 1.22 (1.06-1.41) for non-HF cardiovascular events. CONCLUSION: Glycated haemogobin exhibits a U-shaped association with cardiovascular outcomes in patients with T2D and atherosclerotic vascular disease, with nadir around 7%. CLINICAL TRIAL REGISTRATION: ClinicalTrials.gov Identifier: NCT00790205.