Standard lactations (305 days [d]) of Brown Swiss (BS, n = 54,985), Simmental (SM, n = 114,189) and Holstein (HO, n = 137,703) breeds of cow were analyzed, and their lactation curves were calculated. For the prediction of the milk yield of cows at 305 d and the prediction of lactation – curve standards (LCS), a regression equation system was developed separately for each breed and parities (1, 2, ≥ 3). The regression equation system was used for the estimation of Wood curve parameters for arbitrary milk yield on the basis of the first daily milk recording. An analysis, using the least power norm method (Lp), of the conformity between the calculated Wood's curves and those estimated using the regression equations on the verification data set (n = 34,374) indicated that the equations can provide a good prediction of LCS. Our results indicate that the expected milk yield in a standard lactation and the lactation curve trajectory of an individual cow in a herd may be predicted using a simple algorithm on the basis of the proposed regression equations and known daily milk yield in the initial period of lactation.