Objectives: To increase patients' self-efficacy for initiation of physical activity, there is a need to include physical activity into patient education in clinic settings. In this study, we aimed to assess the effectiveness of multimedia messaging service (MMS) education on exercise self-efficacy in patients with type 2 diabetes mellitus (T2DM). Methods: We used a quasi-experimental, pretest-posttest design to study 98 patients with T2DM. The intervention group received MMS education targeting exercise self-efficacy for 2 months, and the control group received routine care only. Patients in both groups completed the Exercise Self-efficacy scale at 3 stages (at baseline, at 4 weeks, and at 8 weeks post-intervention). Results: We found a slight increase between baseline, first follow-up, and second follow-up in interpersonal and competing demands factors in the intervention group (p = .002, p = .001, respectively), but no improvement in the control group in any of the 3 factors over time (p > .05). Also, Cohen's d values indicated a medium effect size in all exercise self-efficacy subscales (interpersonal [0.734], competing demands [0.665], and internal feelings [0.696]). Conclusions: Health education using theoretically-based MMS targeting exercise self-efficacy was effective and affordable in promoting and changing patients' beliefs and physical activity behaviors.