Calcium phosphate is sparingly soluble at neutral pH but the milk of nearly all species contains many times the concentrations of calcium and phosphate that this solubility allows. The explanation for this apparent paradox is that milk calcium is mainly present in the form of thermodynamically stable nanoclusters formed from casein phosphoproteins and amorphous calcium phosphate. A simplified model system comprising the main milk salts and a hydrophilic N-terminal phosphopeptide from β-casein, at or about neutral pH, forms complexes called calcium phosphate nanoclusters. It is shown that two constants, K S and y, together with the empirical chemical formula of the nanoclusters allow the concentrations of all the important chemical species in this model system to be calculated. The method has been extended successfully to the more complex problem of calculating all the important chemical species in milk. It is suggested, on the basis of a simple analysis of the thermodynamics of the formation of nanoclusters, that their size and stability is determined principally by three factors. These are the relatively low free energy of formation of the amorphous core material, the particular pattern of phosphorylation of seryl residues in the caseins that gives a large chelate effect during the sequestration of the calcium phosphate and the unfolded conformation of the caseins that allows a high density of phosphate centres on the core surface.