In 2003, after over 20 years of minimal health insurance coverage in rural areas, China launched a heavily subsidized voluntary health insurance program for rural residents. The authors use program and household survey data, as well as health facility census data, to analyze factors affecting enrollment into the program and to estimate its impact on households and health facilities. They obtain estimates by combining differences-in-differences with matching methods. The authors find some evidence of lower enrollment rates among poor households, holding other factors constant, and higher enrollment rates among households with chronically sick members. The household and facility data point to the scheme significantly increasing both outpatient and inpatient utilization (by 20-30 percent), but they find no impact on utilization in the poorest decile. For the sample as a whole, the authors find no statistically significant effects on average out-of-pocket spending, but they do find some-albeit weak-evidence of increased catastrophic health spending. For the poorest decile, by contrast, they find that the scheme increased average out-of-pocket spending but reduced the incidence of catastrophic health spending. They find evidence that the program has increased ownership of expensive equipment among central township health centers but had no impact on cost per case.