Despite the recent dramatic rise in student homelessness in the U.S., little research evidence exists on the effects of homelessness programs and interventions on students and young people. This paper examines the effects of a homelessness prevention and rapid rehousing program--which combines temporary rental subsidies with light-touch case management--on homeless and highly mobile students in a large urban school district. By linking detailed district administrative data with programmatic data, I create a novel dataset that allows me to estimate impacts to students in the immediate weeks and months following exposure to the rehousing intervention. Specifically, I use generalized and event study difference-in-differences models to estimate within-student differences in district and school mobility, attendance, and behavioral outcomes before and after beginning participation in the program. I find that the treatment improves student behavior, significantly reducing the likelihood of students having multiple behavioral incidents in a month, but increases students' absence rates and chronic absenteeism, particularly for students who are rehoused outside the city proper but remain enrolled in the central district. These results highlight both the positive impacts of this type of intervention, as well as unintended consequences, raising questions around the priorities and inherent tradeoffs of rehousing programs.