Estimation of surface radiation budget is a crucial step to analyze the climate effects caused by rapid urbanization. This paper reports a study of the integration of remote sensing images and ancillary data for analyzing the spatial and temporal variations of surface radiation budget in Beijing, China. Landsat-5 Thematic Mapper (TM) images and meteorological data of Beijing metropolitan area acquired in the summer and winter were used to calculate land surface parameters and surface radiation fluxes, including shortwave net radiation, effective radiation and net radiation. Validation with in situ measurements shows that the calculation of net radiation yielded high accuracy. It suggests that the integration of remote sensing and ancillary data provide an applicable and feasible routine for analysis of surface radiation budget in urban environment. In order to understand the spatial patterns of surface radiation budgets, parameters, such as radiation fluxes, albedo and land surface temperature, were analyzed in terms of variations among different land cover types. Results indicate that the city can be characterized as a "basin" of net radiation in the summer, while it is characterized as a "plateau" in the winter. The albedo and land surface temperature were two primary factors contributing to the spatial variations of net radiation, while the solar elevation angle controlled the seasonal variations of the absolute amount.