Solar radiation is a very important and major variable in crop simulation models. However, it is measured at a very limited number of meteorological stations worldwide. Models were developed to estimate daily solar radiation in Chongqing, one of the most important agricultural areas in China. Several routinely observed meteorological variables including daily maximum and minimum temperatures, daily mean dew point temperature, fog and rainfall had been obtained, investigated and analyzed from 1986 to 2000 for Chongqing. The monthly mean daily solar radiation at this location ranged from a maximum of 15.082 MJ m(-2) day(-1) in August and a minimum of 3.042 MJ m(-2) day(-1) in December. A newly developed model that included all selected variables proved the best method with a RMSE value of 2.522 MJ m(-2) day(-1). The best performed models for different seasons were further evaluated according to divide-and-conquer principle. The model using all selected variables provided the best estimates of daily solar radiation in winter and autumn with RMSE values of 1.491 and 2.037 MJ m(-2) day(-1), respectively. The method involving temperatures and rainfall information could be used to estimate daily solar radiation in summer with a RMSE value of 3.163 MJ m(-2) day(-1). The model using temperature, rainfall and dew point data performed better than other models in spring with a RMSE value of 2.910 MJ m(-2) day(-1). (C) 2010 Elsevier Ltd. All rights reserved.