The knowledge of the spatial and temporal patterns of Surface Air Temperature (SAT) is essential to monitor a region¿s climate and meteorology, quantify surface exchange processes, improve climatic and meteorological model results, and study health and economic impacts. This work analyzed correlations between SAT and geophysical land surface variables, Land Surface Temperature (LST) mainly, to establish operative techniques to obtain spatially-continuous land SAT maps from satellite data, unlike data provided by meteorological station networks. The correlations were analyzed by using EOS-MODIS images, meteorological station network data, and geographical variables. Linear regressions with MODIS-retrieved LST data gave SAT with uncertainties higher than ±2K during daytime and of ±1.8K at night-time. Nevertheless, SAT uncertainties decreased up to ±1.2K when other satellite-retrieved surface parameters, i.e. vegetation index and albedo, together with meteorological and geographical data were considered as terms of multivariable regressions. The equations finally proposed were shown to work properly for different land covers.