The experimental method is extensively used to determine the temperature of a photovoltaic (PV) module at different hours of a day. In this method, the module temperature is measured using a temperature sensor mounted on the back of PV module. However, the experimental measurements have high cost and are not applicable everywhere. In this study, an optical–thermal model was used to predict all the PV module layer temperatures in two cases: tilted toward the south and fixed on a two-axis sun tracker. The impact of accurate consideration of the wind velocity and the ambient temperature on the PV module temperature was the main strength of the present simulation. This was carried out testing several correlations for prediction of convection heat transfer coefficient in the modeling process. The front and back layer temperatures as well as the silicon (Si) layer temperature of PV module were separately determined. To verify the results of the simulation, the temperatures of four PV modules measured in four different locations of the world, namely, China, Germany, Australia, and Brazil, were used. The results showed that the present study predicts the temperature of PV module more accurately compared to the previous studies. It was also shown that the average temperature errors between the measured and the predicted temperatures relative to the maximum module temperature were 2.19%, 2.3%, and 2.85%, for Australia, Brazil, and Germany, respectively.