Fundamentals of the method of transverse displacements for calculating the HF radio-wave propagation paths are presented. The method is based on the direct variational principle for theoptical path functional, but is not reduced to solving the Euler— Lagrange equations. Instead,the initial guess given by an ordered set of points is transformed successively into a ray path,while its endpoints corresponding to the positions of the transmitter and the receiver are keptfixed throughout the entire iteration process. The results of calculation by the method of transverse displacements are compared with known analytical solutions. The importance of using onlytransverse displacements of the ray path in the optimization procedure is also demonstrated.