Patient education is a key component of diabetes care. Limits in resources often prevent the participation of many patients with type 2 diabetes to structured education programs. The identification of predictors of response to group education could help in selecting those patients in whom the intervention is more cost-effective. A structured interactive group program was proposed to a consecutive series of 150 type 2 diabetes patients, who were then followed prospectively in 24 months, with measurements of HbA1c, BMI, quality of life, eating habits. For comparison, another consecutive series of 113 patients who had received no intervention was also observed for 12 months. A significant reduction in HbA1c was observed in the intervention group at 12 and 24 months (from 7.5 ± 1.4 to 6.9 ± 1.2 and 6.6 ± 1.1% at 12 and 24 months, respectively, both P < 0.01), with no variation in BMI and quality of life. A sustained reduction in total energy, protein, and fat intake was observed after education. The proportion of success (HbA1c < 7% and/or HbA1c reduction from baseline > 1%) in the intervention group was 60.7% (vs. 38.1% in controls) and 63.3% at 12 and 24 months, respectively. In the intervention group, patients with success at 12 months showed lower baseline HbA1c, BMI, duration of diabetes, protein, and cholesterol intake. Patients with a lower duration of diabetes appear to have a greater response to structured group education, whereas age is not a predictor of response. Therefore, educational intervention should be planned in the earlier phases of the disease.