The quantitative estimation of the Fe3+/Fe2+ ratio has a crucial interest to know the upper-mantle oxidation state. The relative proportions of Fe3+ and Fe2+ are usually estimated from electron microprobe analysis (EPMA) assuming chromite stoichiometry. In order to test a new technique to estimate the Fe3+/Fe2+ ratio and to avoid EPMA systematic errors, X-ray photoelectron spectroscopy (XPS) has been used to determine the Fe3+/Fe2+ ratio in four chromite samples from Cuba and the results compared with those obtained by EPMA. Due to the high surface sensitivity of XPS the samples were etched by removing the outermost layers by bombarding with an Ar+ ion gun. Therefore, the surface, as well as 8, 16 and 24 nm in-depth levels, were analyzed, revealing that the spectra change significantly with depth. Thus, from the surface down to 24 nm deep, Fe3+ concentrations decrease from values up to 10 at.% to values below the detection limit of XPS (0.1 at.%). By contrast, a systematic Fe3+ detection (between 1.32 and 4.04 at.%) was observed by EPMA. Consequently, XPS allows us to calculate ferric iron contents, a key datum for the calculation of the real structural formula of chromite. XPS can therefore be considered a useful tool in evaluating mantle oxidation state.