The effect of competition on the quality of health care remains a contested issue. Most empirical estimates rely on inference from non experimental data. In contrast, this paper exploits a pro-competitive policy reform to provide estimates of the impact of competition on hospital outcomes. The English government introduced a policy in 2006 to promote competition between hospitals. Patients were given choice of location for hospital care and provided information on the quality and timeliness of care. Prices, previously negotiated between buyer and seller, were set centrally under a DRG type system. Using this policy to implement a difference-in-differences research design we estimate the impact of the introduction of competition on not only clinical outcomes but also productivity and expenditure. Our data set is large, containing information on approximately 68,000 discharges per year per hospital from 160 hospitals. We find that the effect of competition is to save lives without raising costs. Patients discharged from hospitals located in markets where competition was more feasible were less likely to die, had shorter length of stay and were treated at the same cost.