The adsorption and precipitation of P onto pure calcite led to hypothesize carbonate contributes to P retention in calcareous soils. However, most studies have shown that Al–Fe oxyhydroxides overwhelm the role of carbonate in P adsorption, even in calcareous soils. A previous study of arid region soils with aeolian-derived carbonate from New Mexico supported the hypothesis that the carbonate phase may control adsorption. In batch experiments with CaCl2 as background electrolyte, carbonate was selected as the most contributing variable in regression analyses between soil properties and P sorption. We later suspected that the choice of electrolyte might have biased this result. We found that even diluted CaCl2 enhanced P sorption and the precipitation of calcium phosphate compounds in calcareous soils. In this study we re-examined factors controlling P sorption in arid, pedogenic carbonate-rich soils by: (i) determining if calcite appears to control P adsorption when an electrolyte other than CaCl2 is used, and (ii) correlating microscale concentrations of P and Ca by applying electron microprobe analysis (EMA) to intact soil thin sections. Carbonate was always selected by multiple regression as the most contributing variable, irrespective of the electrolyte choice. Chemical fractionation of bulk samples showed that the largest fraction was HCl-extractable P (Ca-bound P). Microprobe analysis showed that in situ P concentrations were locally higher in carbonate coatings, and that P distribution in microtransects through soil aggregates was positively correlated with that of Ca and F. However, P distribution did not correlate with that of Fe, Al and Si. Results confirmed that accumulation of secondary, aeolian-derived and high-specific surface area, calcium carbonate is the primary geochemical agent capable of fixation and retention of P within the soil profile in this arid environment.