This paper adds to the literature on wealth effects on consumption by disentangling house price effects on consumption for mainland China. In a stochastic modelling framework, the riskiness, rate of increase and persistence of house price movements have different implications for the consumption/housing ratio. We exploit the geographical variation in property prices by using a quarterly city‐level panel data set for the period 1998Q1–2009Q4 and rely on a panel error correction model. Overall, the results suggest a significant long‐run impact of property prices on consumption. They also broadly confirm the predictions from the theoretical model.