Defining optimal grinding regimes with the use of traditional methods of mathematical programming and numerical analysis usually turns out to be not effective enough, therefore, solving this problem on the basis of evolutionary methods of optimization is presented in this paper. Depending on the features of technological process, there may be several optimality criteria, so the problem transforms into multi-objective optimization.Premature convergence of the algorithm, as well as general low level of fitness among the obtained results and significant fluctuations of the average values of fitness for the sequence of generations can obstruct proper definition of the processing parameters. Analysis of studies and publications related to grinding process optimization revealed the problem of configuration of the fundamental evolutionary operators, which remains relevant for the conditions of the applied problem.In order to prevent premature convergence of the algorithm, it is important to provide gradual concentration of the problem solutions set in the direction of the global extremum area. In that case, genetic algorithm parameters should be cus-tomized to provide improvement of the average fitness of population based on the obtained results and simultaneous search of new solutions in the feasible region. Values of the weighting factors of the complex optimality criterion are defined on the basis of configuration of the area of perspective solutions. Results of the technological process of grinding optimization using pre-sented evolutionary algorithm, classical genetic algorithm, and also such evolutionary methods as ant colony optimization method, particle swarm optimization method and scatter search method reveal an advantage of the suggested approach in convergence rate with stable reliability for all the considered algorithms.Thus, taking into account features of the optimal grinding parameters search using evolutionary methods, in this paper recommendations are presented concerning an algorithm for the stated problem resolving and preventing from complications typical for this way of solving.