Extraction of soils with CO2-saturated water is one of the Swiss reference methods for phosphorus and potassium fertilizer recommendations. Laboratory experiments show that a significant role is played by interactions between CO2, calcium, phosphate, the only slightly soluble calcium phosphates and lime. Since soil air exhibits increased CO2 content, these solution equilibria also influence the immediate plant-availability of P in the field. If a soil sample is extracted repeatedly with CO2 water, the extract will always contain similar P concentrations. This reflects the conditions in the field: a soil can supply plants with P over a fairly long period, even if it periodically dries out or becomes waterlogged. The homeostasis of soils as shown in the CO2 method explains why it takes years of over- or underfertilisation for the P supply of the plants to be affected. CO2-K also decreases only gradually during sequential extractions. Consequently, the CO2 method yields a good approximation of the readily plant-available P and K. Although reproducible, the results of the sequential extractions with ammoniumacetate and EDTA (‘AAE10’) are so different from soil to soil that no direct connection with plant availability can be deduced from a chemical perspective.