There are only a few studies that recently have been made for regionalizing the mean temperature in Mexico. Nevertheless, none of them has really reflected the complex climatic variability of the country. A different approach was used in a set of 49 climatological stations containing long-term and high-quality annual Air Surface Mean Temperature (ASMT) data. Principal Component Analysis (PCA) was applied using an oblique-rotated solution (Promax, kappa=2) to an S-mode configuration. The analysis on the annual gradients of the dataset, has successfully extracted six clear groups of stations that vary coherently across time. The ASMT climatic regionalization shows great geographic consistency with the Mexican climatology. The PCA resulting climatic regions also highlights the important influence of the (non-extensively studied, yet) orography of Mexico.