Interventions that combine unconditional permanent housing with support services, known as Housing First approaches, generally improve housing outcomes for people who have experienced chronic homelessness. However, little is known about their long-run outcomes or the consequences of ending such services. We investigate both aspects by examining the long-run effects of an intensive support program on the housing, employment, and health outcomes of chronically homeless people in Australia. Evaluating the three-year program over six-years using a randomised controlled trial, we document substantially higher rates of housing and better employment outcomes during the program period for the treated group, but no substantial changes in health. Three years after the program ends, we observe no significant differences between the treatment and control group with respect to any outcomes, including housing. Our results imply that stable housing is a necessary but not sufficient condition to overcome multiple sources of economic and social disadvantage.