The polar wind, a plasma outflow along open geomagnetic field lines from the Earth's ionosphere, has been the subject of many theoretical studies. Since the proposal of its existence more than 30 years ago, the polar wind has been described by various theoretical modeling approaches. In this paper, we first describe the two traditional modeling approaches for the classical polar wind: moment-based models and collisionless kinetic calculations; and discuss how other models of the outflow were formed by combining these two traditional approaches. We then discuss how kinetic calculations have been improved to take into account the effects of Coulomb collisions, which may contribute to the formation of double-peaked H+ velocity distributions in the polar wind. As observations of escaping O+ fluxes have guided theoreticians to go beyond the classical polar wind description, we discuss how models were modified or constructed in order to take into account the various non-classical effects, including those due to photoelectrons, which may also provide an explanation for the satellite observations of day-night asymmetries in addition to the acceleration of the O+ ions.