Background: policymakers have debated whether test scores represent students’ maximum level of effort, prompting research into whether student-level financial incentives can improve test scores. If cash incentives are shown to improve students’ test performance, there can be questions as to whether test scores obtained in the absence of financial incentives accurately reflect what students know and can do. This can raise concerns as to whether test scores should be used to guide policy decisions. Purpose: this study used meta-analysis to estimate the effect of student-level incentives on test performance. The study also included a narrative review. Research design: twenty-one studies conducted in the united states and internationally were included in the meta-analysis. Effect sizes were estimated separately for mathematics, reading/language arts, and overall achievement. Findings: financial incentives had a significantly positive effect on overall achievement and on mathematics achievement, but no effect on reading/language arts achievement. The narrative review suggested mixed effects with regards to whether treatment estimates could be sustained after the removal of the incentives and whether larger cash payments were associated with stronger program impact. Programs that offered monetary incentives in conjunction with other academic supports tended to show stronger effects than programs that offered incentives alone. Conclusion: the findings raise questions as to whether policymakers should use scores from low-stakes test to inform high-stakes policy decisions. The study cautions against using scores from international assessments to rank order countries’ educational systems or using scores from state achievement tests to sanction schools or award teacher bonuses. 2020 Teachers college, columbia university. All rights reserved.