Land surface temperature (LST) is a key magnitude in many exchange processes between the surface and the atmosphere. LST measurement from satellites provides an efficient way to monitor its change across wide areas on Earth, an essential issue being LST validation using in situ measurements to assess its accuracy and precision. Presently, there are two widely used methodologies: temperature measurements made by wideband radiometers observing the land surface with a given viewing angle and a limited field of view, and measurements provided by total radiation pyrgeometers with a nearly hemispheric field of view. Although both measurements are correlated, they are not equivalent; thus, it is relevant to establish their differences when they are used as ground reference for the thermal infrared sensors. Both methodologies have been compared over a homogeneous grass and a heterogeneous vineyard under different atmospheric conditions. The results show good correspondence in the first case, with differences +0.3±1.0 K, while in the second one the discrepancies can be significant (-1.7±1.1 K) affecting the reliability on the results of the validation of satellite-borne sensors.