At present, ionization energy losses during colli sions of fast charged particles with atoms of a target substance are usually calculated using the Bethe-Bloch formula with standard corrections [1][2][3][4], including the Barkas factor. The need to introduce the latter correction into the Bethe-Bloch theory of energy losses was recognized upon experimental observation [5] of a several percent difference between the ranges of π + and πmesons with identical energies in a photoemulsion. Until the Barkas correction is small, it is possible to calculate this quantity in the sec ond order of perturbation theory [6] in the region of its applicability-i.e., at Z/v Ӷ 1, where Z is the projec tile charge and v is its velocity (here and below, atomic units are employed). However, if the energy losses have to be determined for heavy ions rather than light ones, the Barkas correction can become large (about 100% of the value given by the Bethe-Bloch theory) [1, 7-10], so that a nonperturbative consideration becomes necessary. Unfortunately, exact nonperturbative quan tum mechanical solution of the problem is yet not available. For this reason, the Barkas correction, whenever significant, is calculated using various fitting parameters or rough approximations based on classi cal notions [5,7].Quantum mechanical calculation of the Barkas correction is only possible in simple cases of numerical solution of the Schrödinger equation [11]. This cor rection introduces additional terms into the Bethe-Bloch formula, which depend only on the odd powers of the charge of a retarded particle. For this reason, it is a common practice to use model ion-atom interac tion potentials. For a short range model potential, the exactly calculated effective stopping involves both even and odd powers of the ion charge. Therefore, only terms dependent on the odd powers of this charge represent the Barkas correction, although the Schrödinger equation has still to be numerically solved [12]. In selecting a model potential, its effective radius α has to be introduced. Usually (see, e.g., [10,12]), a Yukawa type potential is selected and it is assumed that α = v/ω, where v is the ion velocity and ω is the char acteristic atomic frequency. This choice of α is based upon the coincidence of energy losses on a short range potential with parameter α and the classical Bohr for mula for energy losses at Z/v 2 Ӷ 1 and ω/v Ӷ 1 [13]. It should be noted that the Yukawa potential is not the only possible short range type. Some other potentials qualitatively reproducing the behavior of the Yukawa potential can also be used to determine the Barkas cor rection without any loss of accuracy. Thus, of all the standard corrections to the Bethe-Bloch formula, despite its wide use, the Barkas correction is still the least studied one and has no commonly accepted rep resentation.In the present work, a model potential has been selected that allows the Barkas correction to be analyt ically determined by quantum mechanical methods and simply calculated without any loss of accuracy ...