Optical spectroscopy has been explored since 70's as a rapid and reliable technique for quality evaluation and detection of irregularities in food products.Recently it is successfully used as a faster, cheaper and eco-friendly alternative for nutrients and fertility analysis of soils, regarding conventional chemical processes.This method allows simultaneously estimated several physicochemical properties at different spectral ranges, however due to the mixed composition, spatial variation and temporal changes of samples, has been necessary the development of stepwise multivariate regression methods to quantify the information contains in the optical spectra.This review presents an overview of the main conceptual and analytical develops in the use of optical spectroscopy for characterization of agricultural samples, beginning with a description of soils as a non-renewable natural resource, the importance and needs of its conservations, as well as the challenge that represents the characterization of highly complex and multicomponent organic systems linked through several chemical interactions.The spectral response of some soils with different properties is presented, followed by the main multivariate statistical methods used in spectral analysis to decouple the physicochemical information contained in the optical signals.Finally, it is shown the hyperspectral imaging technique as a remote sensing method useful to compositional and structural analysis of soils point-to-point in large areas, as well as for the mapping of physicochemical properties in airable lands.