The calcretes of the Kalahari are amongst the thickest in the world representing pedogenic episodes in a semi-arid climate during Pliocene to Recent times. The descriptive morphological terminology of Netterberg is used to describe the calcrete types and a differentiation into simple and composite profiles is made. A pedogenic/diagenetic scheme has been constructed using all available data. Early calcite cementation is induced by two mechanisms. Slow evaporation and/or CO2 loss causes the precipitation of low-Mg calcite, whereas rapid evaporation and/or CO2 loss precipitates predominantly high-Mg calcite, in thermodynamic disequilibrium with the low Mg/Ca ratio vadose water, and minor aragonite (which transforms rapidly to low-Mg calcite). High-Mg calcite is also precipitated from high Mg/Ca ratio vadose waters in calcretes developed on Mg-rich host lithologies and by capillary rise from shallow groundwaters in saline depressions. Calcite precipitation may be passive (cement), displacive or replacive, in the latter released silica migrating down-profile to precipitate length-slow chalcedony, clinoptilolite (saline conditions), length-fast chalcedony and megaquartz (non-saline conditions). Displacive introduction of calcite takes place from highly supersaturated solutions due to rapid evaporation (with CO2 loss) of vadose waters. During low-Mg calcite precipitation (in a ‘closed’ system) the Mg concentration of the resulting solution increases. This, combined with Mg released during high-Mg to low-Mg calcite transformation, induces precipitation of authigenic palygorskite, sepiolite and minor dolomite. Vadose dolomite is often present whilst some dolomite may be precipitated at the mixing-zone of vadose and phreatic waters. The proposed model applies to the Kalahari calcretes, although it may perhaps be extrapolated to other areas. Further detailed studies, involving analyses of pore water chemistry, soil microclimate, and trace element and isotopic analyses of individual cements are necessary.