Seven existing models and four proposed models were calibrated and evaluated to calculate the monthly average daily global solar radiation (GSR) on a horizontal surface in Wuhan city, China, using meteorological data measured from 2006 to 2011. The results show that the sunshine duration (n) is an important parameter for estimating the GSR and adding the dew point temperature (DP) is a significant effect in humid regions especially in Wuhan. Notably, all the tested models that used only the sunshine ratio or the sunshine ratio combined with other parameters exhibited satisfactory estimation performance. Adding the maximum and minimum air temperatures to the sunshine ratio in the Chen and Li model yielded a considerable improvement over the existing models, and adding the DP in the new model resulted in the most accurate estimations of solar radiation. Additionally, using the simple linear Ångström-Prescott model was better than using the more complex Bahel model or the Ododo model. Models based on the maximum, minimum and average air temperature (Li Huashan model) and DP (Li model) exhibited poor performance. The worst performance was displayed by the Badescu model, which uses only cloud cover as a key input. Finally, we propose a strategy for selecting the most appropriate model for estimating the GSR.