We study the long-run impacts of health insurance promotion in Northern Ghana. We randomly provide three overlapping interventions to promote enrollment: subsidy, information campaign, and convenient sign-up option, with follow-up surveys seven months and three years after the initial intervention. Our interventions, especially the subsidy, promote enrollment and healthcare service utilization in the short and long runs. We also find short-run health status improvements, which disappear in the long run. We find suggestive evidence on decreased investment in disease prevention and selection that may help explain this pattern of health status changes.