We examine the impact of removing user fees for healthcare in rural Ghana using data from a randomized experiment that includes rich information on objective measures of child health status. We find that free care increased use of formal healthcare shifting care seeking away from informal providers, with particularly strong effects for children who were anaemic at baseline. There was no health effect on the intervention population taken overall. However, consistent with the utilization findings, there were health improvements amongst those with anaemia initially. Further benefits included a large reduction in health spending, with the effect greater at higher levels of the medical spending distribution. Free care was found to have no influence on a range of malaria prevention behaviours or on the incidence of self-reported illness, suggesting that ex-ante moral hazard is unlikely to be a concern in this particular setting. © 2013 Elsevier B.V.