OBJECTIVES: To compare the effects of adequate and inadequate glycemic control, measured by HbA1c levels, on long-term complications of type 2 diabetes patients of a brazilian healthcare insurance plan, using a computer-based Diabetes Mellitus Model (DDM). METHODS: The DDM predicts short- and long-term complications over ten years using data from previous published studies. Demographic and clinical model parameters were based on 400 type 2 diabetes patients from a private healthcare plan. Two simulations were conducted: in the first scenario, patients stayed on target - HbA1c levels at 7% - during 10 years (adequate glycemic control); in the second scenario, patients stayed on baseline values and had a 2% increase in HbA1c levels during the 10-year period (inadequate glycemic control). Scenarios were then compared in respect to occurrence of disease complications. RESULTS: The relative risks (RR, simulations 1/2) were 0.87, 0.80, 0.71, and 0.90, respectively, for the occurrence of ophthalmic, renal, neural and vascular, and cardiovascular complications. Simulation 1 had a mortality risk reduction of 11%, compared to simulation 2. The RR ratios for specific events, such as photocoagulation, cataract extraction, macular edema, dialysis, amputation, myocardial infarction, and heart failure were: 0.96, 0.82, 0.61, 0.77, 0.88, 0.86, 0.76, respectively. CONCLUSION: Although there is a need to corroborate the results of these simulations with real, long-term clinical data, the computer-based model estimates have shown lower complication rates for all events evaluated and a reduction in mortality risk in adequate glycemic control group, comparatively to inadequate glycemic control group, using a 10-year time horizon, in a brazilian private healthcare insurance population.