This paper is aimed at improving the system of assessing the breeding value of bulls by set complex indicators of their daughters’ lactation performance. The authors carried out studies using a database for first-calving cows of black-motley population of the Moscow region for the period 2011-2015. The authors used a previously optimized structure as the initial equation of the BLUP mixed model for assessing the genetic features of bull sires for each indicator of their daughters ’ lactation performance. This model takes into account and eliminates the paratypic effects on the variability of studied parameter to the greatest extent. To construct a selection index in the studied population, coefficients of heritability, as well as phenotypic and genetic variants (covariants) of traits were calculated. As a result, the form and weight coefficients for each trait in the selection index were determined and formalized. To determine the effectiveness of using the developed selection index, the breeding process in the population was simulated during the formation of animal breeding groups based on the different intensity of their selection. The results obtained convincingly prove the advantage of the selection index in assessing the genetic values of animals in terms of the totality of the analyzed traits and in their selection into breeding groups as the birth fathers of bulls and cows in the current population. The advantage of selecting animals according to the index was 3.0-10.0% as compared with the selection only for milk yield, milk fat and milk protein, and 1.7-30.9% for the fat and protein content. Thus, the study results convincingly prove the fundamental necessity to use a selection index when assessing animals (comprehensive assessment of breeding and productive features of individuals) and optimizing breeding programs for genetic improvement of dairy cattle breeds and populations in Russia.