Background: Because medication adherence is linked to better diabetes outcomes, numerous interventions have aimed to improve adherence. However, suboptimal adherence persists and necessitate continued research into intervention strategies. This study evaluated the effectiveness of an intervention that combined storytelling and peer support to improve medication adherence and health outcomes in adults with diabetes. Methods: Living Well with Diabetes was a cluster randomized controlled trial. Intervention participants received a six-month, 11-session peer-delivered behavioral diabetes self-care program over the phone. Control participants received a self-paced general health program. Outcomes were changes in medication adherence and physiologic measures (hemoglobin A1c, systolic blood pressure, low-density lipoprotein cholesterol, body mass index). Results: Of the 403 participants with follow-up data, mean age was 57 (±SD 11), 78% were female, 91% were African American, 56.4% had high school education or less, and 70% had an annual income of < $20,000. At follow-up, compared to controls, intervention participants had greater improvement in medication adherence (β = −0.25 [95% CI -0.35, −0.15]). Physiologic measures did not change significantly in either group. Intervention participants had significant improvements in beliefs about the necessity of medications (β = 0.87 [95% CI 0.27, 1.47]) concerns about the negative effects of medication (β = −0.91 [95% CI -1.35, −0.47]), and beliefs that medications are harmful (β = −0.50 [95% CI -0.89, −0.10]). In addition, medication use self-efficacy significantly improved in intervention participants (β = 1.0 [95% CI 0.23, 1.76]). 473 individuals were enrolled in the study and randomized. Discussion: Living Well intervention resulted in improved medication adherence, medication beliefs, and medication use self-efficacy but not improved risk factor levels.