The sky thermal irradiance (R-sky) is an important factor in meteorological, climatic, and energy balance studies. There are several empirical models for predicting R-sky; however, these are based on the meteorological parameters measured in non-arid climatic conditions and an empirical model for predicting R-sky in arid climates is unavailable. Therefore, this study presents two empirical models for predicting the R-sky under clear sky conditions, common in arid regions such as the Arabian Peninsula. One is based on the sky equivalent emissivity (epsilon(sky)-model) and the other on the sky equivalent temperature (T-sky-model). Meteorological parameters (i.e., air temperature, relative humidity, solar radiation, and downward R-sky) were measured over 3 years and used to develop and validate the proposed models. The results showed that the proposed epsilon(sky)- and T-sky-models had the highest predictive accuracy with a coefficient of determination of 0.86 compared with previously published models. In addition, the proposed epsilon(sky)- and T-sky-models resulted in low values of root mean square and mean absolute percentage errors (7.2 and 7.8 W center dot m(-2) and 1, 2 %, respectively); they were able to precisely predict R-sky at any time in a clear arid climate compared with the prohibitive and time-consuming measurements.