OBJECTIVE: The associations of discontinuation of the study medication on major outcomes were assessed in the Action in Diabetes and Vascular Disease: Preterax and Diamicron MR Controlled Evaluation Trial. METHODS: ADVANCE was a factorial randomized controlled trial of blood pressure lowering (a fixed combination of perindopril and indapamide vs. placebo) and intensive glucose control (vs. standard glucose control) in patients with type 2 diabetes. Patients who permanently discontinued the randomized blood pressure-lowering medication during the study period (n = 1557) were compared with others (n = 9583). Cox's proportional hazards models were used to estimate the effects of the discontinuation on the risks of macrovascular events, microvascular events together and separately and all-cause mortality, using discontinuation as a time-dependent covariate. RESULTS: In multivariable analyses, discontinuation was associated with increased risks of combined macro and microvascular events (hazard ratio 2.24, 95% CI 1.96-2.57), macrovascular events (3.23, 2.75-3.79), microvascular events (1.38, 1.11-1.71), and all-cause mortality (7.99, 6.92-9.21) compared to continuing administration of randomized medications during the trial period, which were highest in the first year after discontinuation. These associations were similar in active and placebo groups, except in the first year after discontinuation during which event rates were lower in the active group than in the placebo group (P ≤ 0.01). CONCLUSION: Discontinuation of study medication is a potent risk marker for identifying high-risk patients. Thus it is important that clinicians seek to identify such patients early after discontinuation of treatment. Although some short-term residual effects of previous active treatment can be expected, patients who discontinue require further urgent investigation and management.