Non-adherence to antiepileptic drugs (AEDs) is a common problem (i.e., 58% of patients have some level of non-adherence) for youth with epilepsy, with potentially devastating consequences. Adolescents with epilepsy represent a particularly vulnerable group, given their increased independence, decreased parental supervision, higher risk for deficits in organization and memory, busy and changing schedules, low motivation, and increased susceptibility to peer influence. Existing adherence interventions in epilepsy are not designed to meet the unique challenges faced by adolescents, and there are no efficacious interventions for adolescents with epilepsy. Not surprisingly, without efficacious interventions, adherence worsens during adolescence, further increasing the risk of poor health outcomes during this developmental period. While reminder strategies (e.g., automated digital reminders) are effective for the most common adherence barriers of forgetting and busy schedules, they are likely to be ineffective in increasing motivation. Leveraging social norms comparison methods (i.e., feedback about someone else's behavior related to one's own behavior) offers an opportunity to capitalize on the increased importance of peer influence while simultaneously targeting the low motivation characteristic of adolescents. Recent data in adolescents indicates that social norms interventions have incremental value and improve health behaviors above and beyond standard feedback without peer comparisons. Consistent with the ORBIT model for behavioral intervention development, our aims are to: 1) develop a feasible, accessible, and acceptable mHealth social norms intervention for improving AED adherence in adolescents with epilepsy and 2) obtain preliminary efficacy data and effect sizes for a future clinical trial. We will address these aims through an iterative process, including an adolescent focus group (ORBIT Phase 1a: Define; n=8), extended usage test (ORBIT Phase 1b: Refine; n=10) and a pilot RCT of the final mHealth social norms intervention (ORBIT Phase II: Pilot; n=138). Adolescents with epilepsy who demonstrate non-adherence (< 95% adherence based on PI's previous RCTs; 58% of sample) during baseline will be randomized to either 1) mHealth social norms (automated digital reminders, individualized adherence feedback, and social norms feedback) or 2) control (automated digital reminders and individualized adherence feedback). Both groups will receive active intervention for five months. Primary (i.e., electronically-monitored adherence) and secondary outcomes (i.e., seizure severity, HRQOL) will be assessed post-treatment and 3 months later, respectively. If successful, the results of this study would have a large impact on pediatric epilepsy, with the potential to change clinical practice for treating non-adherence by reducing common barriers to behavioral health care. Because minimal clinician time is required, our mHealth social norms intervention also has potential for sustainability and broad dissemination for epilepsy and other pediatric conditions