Increased attention to the needs of adolescent girls has led to a growing number of programs in low- and middle-income countries (LMICs). Questions remain, however, about what aspects of program design are most effective. This hinders efforts to effectively allocate resources, scale-up programs, and replicate results across settings. To address these issues, we conducted a systematic review to identify lessons learned and gaps in the evidence base. Key findings in the report include: (1) Multicomponent programs tended to outperform single-component programs, although few studies controlled for duration of exposure to the intervention. There were too few studies to draw conclusions about the durability of program effects over time; (2) Longer program exposure improved program effectiveness and may contribute to the durability of program effects over time, although selection bias was a limitation in a number of studies; (3) Evidence on whether multilevel interventions enhance benefits for girls relative to single-level interventions was inconclusive. No identified studies assessed the relative effect of booster "add-ons" (versus no booster) or varied the saturation level of the program in communities; and (4) Few evaluations of girl-centered programs to date have rigorously addressed implementation science questions, highlighting a robust and urgent research agenda for the future. [Additional support by the Nike Foundation.]