Over the course of just two years, at least six reviews have examined interventions that seek to improve learning outcomes in developing countries. Although the reviews ostensibly have the same objective, they reach sometimes starkly different conclusions. The first objective of this paper is to identify why reviews diverge in their conclusions and how future reviews can be more effective. The second objective is to identify areas of overlap in the recommendations of existing reviews of what works to improve learning. This paper demonstrates that divergence in the recommendations of learning reviews is largely driven by differences in the samples of research incorporated in each review. Of 229 studies with student learning results, the most inclusive review incorporates less than half of the total studies. Across the reviews, two classes of programs are recommended with some consistency. Pedagogical interventions that tailor teaching to student learning levels-either teacher-led or facilitated by adaptive learning software-are effective at improving student test scores, as are individualized, repeated teacher training interventions often associated with a specific task or tool. Future reviews will be most useful if they combine narrative review with meta-analysis, conduct more exhaustive searches, and maintain low aggregation of intervention categories.