Abstract: In this paper, parameters of Weibull distribution were used to describe effects of urbanization on surface wind speed. Theoretical analysis indicated that urbanization could lead to a decrease in the scale parameter c and an increase in the location parameter r. Surface weather timing mode observations collected at Beijing and other 8 cities during 1982-2005 were divided into 2 segments based on different time periods and parameters of distribution curve were fitted respectively. Fitting results confirmed the conclusions that with the development of the city, parameter c decreased and parameter r increased in varying degrees, and their probability density peaks both increased. Characteristics of wind speed probability density function curves in different cities could be different. According to this feature, city wind speed probability distribution changes in the 8 cities could be divided into Type A and Type B approximately. The main difference in the features of the curves between the two types was that for Type A, both the strong and calm wind probabilities decreased as shown in both sides of the curve and the distribution become more concentrated; for Type B, the strong wind probability decreased but the calm wind probability increased and the curve moved to the left；r parameter increased by 46.3% in average in Type A cities, which was greater than that in Type B cities by 15.7%. Initial data analysis also indicated that the variation degree of wind speed parameter c and parameter r was related to the speed of urban expansion.