Thermodynamic consistency within the density-wave theory of freezing is briefly discussed. The non equivalence of different routes to calculate the change in a thermodynamic quantity upon freezing is related to the approximations involved in practical calculations. In particular, a new way of calculating entropy and volume changes is proposed. It is shown that this new route avoids an unphysical feature of the entropy formula previously proposed by other authors. The freezing of hard spheres is discussed as an illustration.