The purpose of our study was twofold: (1) to review the literature on high-stakes accountability testing and English learners (ELs) in the USA, applying qualitative systematic review methodology, and (2) to draw substantive conclusions about the impact of high-stakes accountability testing on ELs, reported in studies published between 2001 and 2016. Thirty-seven studies (n = 37) meeting the eligibility criteria were evaluated using the Methodological Quality Questionnaire. Findings indicate construct irrelevant variance (non-random factors that systematically affect ELs' test performance) challenges the appropriateness of inferences drawn about ELs' content knowledge based on assessment performance. Further, language policy in schools, driven by high-stakes tests, promotes increased instructional time for basic skills tutorials with fewer opportunities for problem solving and other higher-order learning activities. Moreover, theory utilization was rare (n = 8, 22%). Among studies that utilized theory, few explicitly referenced it to discuss study findings.