Investigations of X-ray photoelectron spectra from solid samples need corrections for the surface charging effect. For powder samples such as heterogeneous catalysts and their supports, the C(C,H) component of the C 1s peak is often used as an internal standard for the calibration of the binding energy scale. Although this method is widely recognized as suitable for the study of heterogeneous catalysts, we show that a significant calibration bias can be encountered upon comparing samples with different bulk composition. In this paper, a series of SiO2–Al2O3 supports and Pd/SiO2–Al2O3 catalysts with various Si/Al ratios were studied. The spectra issued from these samples were processed with the classical calibration method on the basis of the carbon peak. Important discrepancies in the relative position of the photoelectron peaks were noticed. After systematically discarding instrument-related issues, a true chemical influence of the bulk matrix on the analyzed surface species was evidenced. The extent of this chemical effect was dependent on the composition of the sample and more precisely on its ionicity. Two possible mechanisms for this chemical effect were proposed and discussed. Finally, an alternative calibration method was offered.