The effect of casein on the quality of ice cream with different fat content was studied. According to functional and technological characteristics, micellar casein was selected for the enrichment of ice cream. Using mathematical modeling in the environment of the MathCad-15 package, the mass fraction of micellar casein in the composition of ice cream with a fat content of 0 to 15 % was optimized in order to obtain a high-quality product. At the first stage, the response surface methodology was used to optimize the response functions (overrun, melting resistance, organoleptic characteristics) for the varied fat and protein content. In the second stage, a comprehensive quality score of ice cream was used for modeling as a function of estimates of individual quality indicators, converted into scalable values using weights. The inverse relationship between the values of the optimal protein content and the fat content of ice cream was determined. To achieve the maximum technological effect, in the composition of ice cream with a fat content of 0-5 %, 6-10 % and 11-15 %, the need for micellar casein is 6-5 %, 4-3 % and 2.5-1 %, respectively. According to the results of calculating the percentage of energy value introduced by total protein (more than 20 %), it was concluded that ice cream with a fat content of 0-5 % with mass fractions of micellar casein of 6-5 % and total protein of 9.7-8.7 % can be attributed to the category of products with high protein content. Ice cream with a fat content of 10-15 % with mass fractions of casein micellar of 3-1 % and total protein of 6.7-4.7 % can be attributed to a product with high protein content. The results of the study allow expanding the range of protein-containing ice cream to meet the needs of consumers of different groups