Absenteeism is a prevailing concern in American education, and students experiencing homelessness are particularly vulnerable to high rates of school absenteeism. Despite this increased risk, we find no research in which the primary focus is assessing the efficacy of shelter-based programs that seek to reduce absenteeism among homeless children. Thus, we evaluate the Attendance Matters program, which sought to improve school attendance among homeless students in New York City shelters through interagency coordination, leveraging data to target scarce program resources, and employing evidence-based social work practices. We use administrative data in a quasi-experimental study to evaluate the program’s effects on school attendance and, secondarily, on outcomes of proficiency and stability. Findings suggest that the program resulted in reductions in days absent and the absence rate among K-8 students, though findings for secondary outcomes and attendance outcomes for high school students were inconsistent across model specifications. Results, which likely understate actual program effects, have implications beyond this setting, as they suggest that a low-budget program leveraging evidence-based practices and existing resources can impact this seemingly intractable problem. Education and homelessness policymakers should seek opportunities to test replication in additional settings.