The reliable estimation of the radiative forcing and trends in radiation requires very accurate measurements of global and diffuse solar irradiance at the earth's surface. To improve measurement accuracy, error sources such as the pyranometer thermal offset should be thoroughly evaluated. This study focuses on the measurement and analysis of this effect in a widely used type of pyranometer. For this aim, a methodology based on capping the pyranometer has been used and different criteria for determining the thermal offset have been applied and compared. The thermal offset of unventilated pyranometers for global and diffuse irradiance has been measured under a wide range of cloud, ambient temperature, wind speed, and radiation conditions. Significant differences in absolute values and variability have been observed between daytime and nighttime, advising against correcting the thermal offset effect based only on nighttime values. Notable differences in the thermal offset between cloudy and cloud-free conditions have been also observed. The main results show that the ambient temperature, the radiation, and its direct/diffuse partitioning are the variables more related to the daytime thermal offset.