To analyze long-term consequences of homelessness, the authors compared 388 formerly homeless children 55 months after shelter entry with 382 housed peers, birth to 17, using mother- and child-reported health, mental health, community involvement, cognitive performance, and educational records. Both groups scored below cognitive and achievement norms. Small group differences favored housed 4- to 6-year-olds on cognition and 4- to 10-year-olds on mental health only. Child care and recent stressful events, which were high, were as or more important than prior homelessness. Only children living with mothers were included, potentially biasing results. Policy implications are discussed.