The objective of the paper is an evaluation of optimization technique based on genetic algorithm, concerning an applicability of the method to the inverse analysis. The general principles of the inverse analysis are discussed in the paper and short description of the direct model based on the finite element solution is given. Genetic algorithm is presented next and an implementation of the method into the inverse analysis is shown. Practical application of the algorithm is investigated for copper rings compressed on the Gleeble 3800 simulator. Load-displacement data and shape of the ring after compression were used as input to the inverse analysis. Three optimization methods are compared in this analysis: genetic algorithm, Hooke-Jeeves and simplex. The parameters of the analysis were selected taking into account a similar number of callings of the finite element solver for all methods. Comparison of the results has shown that genetic algorithm is an efficient optimization technique for the inverse method applications. It confirmed good accuracy and convergence as well as avoiding of local minima during the optimization process.