This paper examines the effect of one of the world’s largest social assistance programs, the Minimum Living Security System (MLSS) in China, on its recipients’ smoking and drinking behaviors. In order to address the endogeneity of risky health behaviors and income as well as the issue of selection bias, this paper uses a propensity score matching (PSM) method on the 2012 wave of the China Family Panel Studies Survey (CFPS) to explore the effect. The results indicate that the MLSS receipt decreases the probability of smoking and smokers’ consumption on cigarettes while it does not affect smoking intensity. Meanwhile, there is no evidence that the MLSS has an effect on recipients’ drinking behaviors.