We use soil developmental criteria to interpret carbonate-paleosol genesis in the Plio-Pleistocene St. David Formation, southeastern Arizona. Our study of six paleosol sections indicates that pedogenesis varied during the time the St. David Formation was deposited as well as in different depositional environments. Paleosol features range from those common to modern, well-drained surface soils requiring long periods of stability to those indicating hydromorphic-soil processes and cumulic pedogenesis. Differentiating these pedogenic processes helps to interpret carbon- and oxygen-isotope data. Vadose-zone paleosols reflect paleoenvironment better than hydromorphic paleosols because the likelihood of detrital carbonate and ground-water contamination is less. The morphology and texture of paleosols that formed above the influence of the ground-water table (vadose) are similar to those of nongravelly, well-drained Quaternary soils developed in arid regions. Carbonate-enriched zones are beneath clay-rich, carbonate-depleted zones. Carbonate is disseminated in the matrix and is segregated into small (<= 1 cm diameter) nodules, filaments, and seams; well-developed carbonates are massive. Layers of increased clay accumulation have a higher ratio of fine-to-coarse (or to total) clay than overlying or underlying layers, indicating that clay in these layers was probably translocated. Paleosols that formed within the influence of the ground-water table (hydromorphic) have carbonate and clay accumulations unlike those found in modern, well-drained surface soils. Carbonate is segregated into large (2-10 cm diameter) subspherical masses (glaebules) in a noncalcareous or weakly calcareous matrix. Layers of increased clay accumulation seldom have a genetic relation to zones of increased carbonate. These clay-rich layers lie above or below or correspond to the zones of carbonate aggregates.