Common mental health problems impose significant costs on individuals and societies, yet healthcare systems often overlook them. We provide the first causal evidence on the effectiveness of a pioneering, nationwide mental health service for treating depression and anxiety disorders in England using non-experimental data and methods. We exploit variations in waiting times to identify treatment effects, based on a novel dataset of over one million patients that well represent the English population. Our findings show that treatment improved mental health and reduced impairment in work and social life. We also provide suggestive evidence of enhanced employment. However, effects vary across patients, services, and areas. The programme is cost-effective and provides a blueprint for treating mental health in other countries.