Background: Epilepsy is the common neurological disorder in the world, affecting approximately 50 million people. Anti-epileptic medication non-adherence can be a reason for long term hospitalization, repeated emergency seizure attacks, increased health care cost and frequent absence of work due to poor seizure control. Existed studies of anti-epileptic medication non-adherence in Ethiopia have reported great discrepant and inconsistent results which calls a growing demand of systematic review and meta-analysis. Therefore, this review aimed to show the pooled prevalence of anti-epileptic medication non-adherence among people with epilepsy attending outpatient department. Methods: Literatures were searched from the PubMed/Medline, Science Direct, PsycINFO, Hinnari and Google Scholar for grey literatures. The data were extracted using a prepared data extraction Microsoft Excel format. The data were analyzed using STATA-version 14 (software). The I 2 test was used to check the heterogeneity between primary studies with a corresponding 95% confidence interval. Results: A total of fourteen primary studies of anti-epileptic medication non-adherence were included in the review showing the pooled prevalence of anti-epileptic medication non-adherence to be 39.77 (95% CI: 32.44, 47.10). The highest prevalence [44.13 95% CI: (29.92, 58.34)] was observed among studies used both self-report and medical record review together, and studies used only self-report to screen medication none adherence had the lowest prevalence [37.95% (24.50, 51.41)]. Presence of co-morbid illness [2.27 (95%CI: 1.01, 5.12)], medication side effects [1.84(95% CI: 1.43, 2.38)], substance use or drug abuse [2.01(95% CI: 1.27, 3.20)] had statistically significant association with anti-epileptic medication non-adherence. Conclusion: In this review, we found that there is a high burden of anti-epileptic medication non-adherence among people with epilepsy in Ethiopia. This demonstrates a need for clinicians to give more attention for the monitoring and evaluation of anti-epileptic medication adherence in the health care service. We also highly recommended for the adoption of a standardized and contextualized adherence screening tools.