The article presents an original use of experimental methods for the identification of the regression function in the form of a surface response to the optimization of parameters of ball-cratering method. This article suggests a methodology to determine the parameters' values (load, speed of rotation, minimum and maximum distance) for the ball-cratering method developed for determining the abrasive wear coefficient K c evaluating resistance of the tested coating to abrasion. The paper presents an experimental identification system, developed according to the central composition plan enabling the identification of the regression function in the form of a quadratic form. The objects of identification were the functions of the maximum relative error for diameters of craters' wear and the thickness of the wear ring, based on the values of the ball-cratering method. Analysis of the results of identification experiments were shown. Moreover, the construction and evaluation of the regression function and the use of the utility function to optimize parameters of the ball-cratering method were described. The verification of the developed methodology was performed based on the example of the selection of parameters for the ball-cratering method for a coating with high resistance to heat (AlTiCrN) which was deposited with the use of the physical vapor deposition method. To verify the obtained optimal parameter values in relation to the selected anti-wear coating, wear tests were performed according to the set parameters and were subjected to empirical verification confirming the obtained results. Application of response surface optimization method allows for the adaptation of the method to evaluate various types of coatings. Shortening the testing time is a significant achievement for the researchers.