This paper examines the long-run convergence of houses price across the eighteen Beijing districts using monthly dataset that span from January 2006 to December 2014. Following a pair-wise approach, we conduct unit root tests on all N(N-1)/2 possible pairs of housing price differentials across the N districts of Beijing. By doing so we do not select any base-district or regional average as the benchmark. A two-stage approach is employed . In the first stage, unit root tests are employed to investigate the convergence (stationarity) of house price differentials across Beijing. Over half of the intra-city house price differentials are found to be stationary. The second stage of the investigates the drivers of convergence. The probability of a pair being stationarity is affected by income differentials across the eighteen districts, as well as the demographics differentials and supply-side factors. Last we reveal that the half-life of a shock towards long-run price equilibrium is positively affected by distance and housing supply while little evidence can be found for the influence of income and population density.