Using 11 years of county-level panel data, fixed effect models are estimated to evaluate the impact of the Safe Motherhood (SM) Program in China. Propensity score matching is used to select comparable factual and counterfactual counties. Out of 2013 counties in China, 283 are selected for the treatment group and 1051 for the control group. The results support the causal relationship between the program and its targeted outcomes and the partial effects increases as years of exposure in the program: 7 years' treatment of the program increases hospital delivery rate by 3.992 per 100 live births and decreases Maternal mortality ratio (MMR) due to hemorrhage by 10.229 per 100 000 live births. Further modeling supports the conclusion that the program reduces MMR by enhancing MCH care. With an average annual incremental unit cost for the program of about 318.0 thousand RMB (39.8 thousand USD) per county, we conclude that the SM Program is effective in reducing MMR through the enhancement of hospital delivery. Copyright © 2010 John Wiley & Sons, Ltd.