Housing purchasing power measuring is an important issue for the housing guarantee decision making, which has been studied in the static way. However, the consumption structure and income are not changeless. Thus the dynamic model for urban housing purchasing power is proposed and the main idea is as follows. First, construct the income predicting model with the previous income by the least square method. Second, construct the consumption structure prediction model with differential function. Third, the model for urban resident housing purchasing power measuring has been proposed based on the hypothesis of random walking. At last demonstration analysis has been studied with the data set in Beijing from the year 1990 to the year 2005. And the differentiation between static model and dynamic model has been discussed.