In recent years, China has uniquely implemented various policies to control housing prices, particularly its property-purchasing limitation policy. This research proposes a vector autoregression (VAR) model with likelihood-ratio (LR) tests to examine the effects of such a policy on housing prices at the national, provincial and city levels in China, with the use of monthly data from 2002 to 2013. The results show that at the national level, the effect of the policy is very significant, and the impact on housing prices is far greater than monetary and credit policies. However, the policy is not applicable at the provincial level. The policy has a significant role at the city level in first-tier cities, but no significant effect in second-tier cities. Overall, property-purchasing limitations inhibit the growth of housing prices to some extent, and the effects show strong regional characteristics, especially at the city level. Policymakers should therefore take into account regional characteristics in the formulation and implementation of a property-purchasing limitation policy.