In this article are considered the results of usage of genetic algorithms in an estimation of the parameters of distribution with a bounded scattering region problem. The generalized normal distribution law with a bounded scattering region was studied. This distribution law model is based on the imaginary sources model. It is shown that normal distribution with a bounded scattering region is a five-parameter distribution. The description of a technique of genetic algorithms usage in an estimation of the parameters of five-parameter probability distribution function problem is given. In this article are defined the parameters of genetic algorithm which are necessary for a problem solving. The argumentation of a choice of a distribution parameters searching area is given. This article gives an analysis of an influence of a most commonly used genetic algorithm tunes upon distribution parameters estimation accuracy. As the result of this research the technique of the parameters estimation of a generalized normal distribution is proposed. The technique is based on genetic algorithms. The confirmation of technique efficiency is received. This article provides estimation of calculation accuracy of the distribution parameters and integral index which characterizes estimation accuracy in common.