This paper presents evidence on the effects of access to domestic and international markets on per capita consumption of households using data from rural China. The econometric analysis uses alternative identification schemes to address the potential endogeneity of access to markets. We use straight-line distances to coastline and navigable river, along with the topography of the intervening counties, as sources of exogeneous variations. We also use identification through heteroskedasticity, which does not rely on standard exclusion restrictions. The results from alternative identification schemes show that better access to both domestic and international markets has positive effects on per capita consumption, the domestic market effect is significantly larger in magnitude, and there is complementarity between the access to domestic and international markets.