Despite increasing rate of diabetes, no standard self-management education protocol has been developed in Iran. We designed Persian Diabetes Self-Management Education (PDSME) program using intervention mapping. Effectiveness of program was assessed in newly diagnosed people with type 2 diabetes and those who had received little self-management education. Individuals aged 18 and older (n = 350) were recruited in this prospective controlled trial during 2009-2011 in Tehran, Iran. Patients were excluded if they were pregnant, were housebound or had reduced cognitive ability. Participants were randomly allocated in intervention and control groups. PDSME patients attended eight workshops over 4-week period following two follow-up sessions. Validated questionnaires assessed cognitive outcomes at baseline, 2 and 8 weeks. HbA1c was assessed before and 18-21 months after intervention in both groups. The CONSORT statement was adhered to where possible. A total of 280 individuals (80%) attended the program. By 18-21 months, the PDSME group showed significant improvements in mean HbA1c (−1.1 versus +0.2%, p =0.008, repeated measure ANOVA (RMA)). Diabetes knowledge improved more in PDSME patients treated with oral antidiabetic agents than in those receiving usual care over time (RMA, F = 67.08, p < 0.001). Statistically significant improvements were seen in PDSME patients for self-care behaviors, health beliefs, attitudes toward diabetes, stigma, self-efficacy and patient satisfaction. PDSME program was effective in improving self-management cognitive and clinical outcomes. Results support use of intervention mapping for planning effective interventions. Given the large number of people with diabetes and lack of affordable diabetes education, PDSME deserves consideration for implementation.