The transition from “rural China” to “urban and rural China” has triggered the changes of the occurrence level and vertical extension of peasant households’ social capital, but the heterogeneity of the impact of traditional and new social capital on poverty vulnerability has not attracted attention. Based on this, this paper divides rural household social capital into close, connected and bridging social capital, and puts forward the research hypothesis of different types of social capital have different effects on poverty vulnerability. Based on the data of the China Family Panal Studies (CFPS), the hypotheses of this study were tested from the perspectives of influence effect, transmission mechanism, and heterogeneous effect between groups. The findings are as follows: Firstly, different types of social capital can alleviate poverty vulnerability, but on the whole, the connected social capital formed by horizontal expansion has greater effect than traditional tight social capital, and bridging social capital formed by vertical extension has the least effect. Secondly, different types of social capital mainly affect the poverty vulnerability of peasant household through two mechanisms: risk mitigation and income growth, but the mediating effect of connected social capital is higher (up to 85. 87%). Thirdly, different groups benefit unequally from different types of social capital in alleviating poverty vulnerability. The effect of tight social capital on low vulnerable farmers, connected social capital on high vulnerable farmers and bridge social capital on poor farmers (including absolute poverty and relative poverty) is not significant.
