Intelligent methods are used more and more nowadays, in various fields, whose main purpose is to find better solutions to problems that are considered difficult. Identifying the parameters of transformer winding is a crucial problem. In this study, the authors shall use the 'particle swarm optimisation' (PSO) method to identify the parameters of highfrequency transformer winding, for diagnostic purposes. They validate the approach by comparing their results with the experimental ones. Two cases study of windings with air core were considered. Not only the effectiveness and accuracy of the proposed method is verified, the PSO method is compared with another methods referred to as transfer function and genetic algorithm. The results are in total agreement and show the reliability and the robustness of the PSO method. 2 Objective and principal of the study The primary goal in this study is to accurately identify the transformer winding parameters, starting from FRA measurements.