The theoretical basis for the measurement of nutrient potential is examined and its application to soils is discussed. It is suggested that nutrient potential should be measured in standardized units by defining chemical potential relative to a given standard state, such as the usual standard state of solution chemistry. The unit for all nutrient potentials would then be calories per equivalent. The nutrient capacity is regarded as the ability of a soil to resist change in potential. It is therefore indicated by the shape of the curve relating potential to the quantity of the nutrient in the solid phase. For potassium the relationship between potential and exchangeable potassium was found to be curvilinear, so that each decrement of potassium makes the next decrement more difficult to remove. Curves for five soils differed from each other. This was not only because of differing cation exchange capacities but also because of differences among the types of colloids. The relationship between phosphorus potential and phosphorus sorption was also found to be curvilinear and the slopes for the various soils again differed from each other. During equilibration with phosphate, microbial activity caused biological immobilization of phosphorus. This could be controlled by adding mercuric chloride or chloroform, but toluene was less effective.