This paper evaluates the accuracy and applicability of thirteen empirical global solar radiation models for warm subhumid regions that use daily meteorological data at six stations in Mexico (Yucatan Peninsula). One of the models investigated is a new empirical model developed for estimation of daily global solar radiation from a horizontal surface. The models were evaluated using root mean squared error (RMSE), mean bias error (MBE), mean percentage error (MPE), mean absolute percentage error (MAPE), mean absolute bias error (MABE) and coefficient of determination (R-2). Results show that the newly developed model performs best for all stations; models based on temperature, rainfall and air humidity performed better than models that used temperature data only. If only temperature data are available for estimation of global solar radiation, the Bristow and Campbell model had the best results. The new proposed model can be used in hydrologic and agricultural studies to estimate global solar radiation in warm subhumid regions when temperature, precipitation and humidity data are available. (C) 2015 Elsevier Ltd. All rights reserved.