The problematic nature of the calculation of phase equilibria involving solid components is discussed. The concept of ’the stability criterion’, as introduced, accounts for the necessity of a suitable set of parameters for any given activity coefficient model, yielding an appropriate shape of the Gibbs energy function in the whole domain of the equilibrium diagram. The use of a unique reference state (pure liquid at the temperature and pressure of the system), for both solid and liquid phases, is another fact that must be remembered when correlating and predicting these types of equilibria.