AbstractTwo schemes of the crystalline potential for the slow electron in the conduction band of alkalihalides are proposed. The first one (semiempirical) is based on the assumption that in the case of an ion in the crystal one can use the same energy dependence of logarithmic derivatives (of radial solutions at the atomic sphere radius) as in the case of the isolated ion. The modifications required consist in some potential corrections, that are due to the Madelung field as well as to the relaxation and polarization response for both, hole and electron. The second model (ab initio) suggests the straightforward solution of Poisson's equation. Both approaches are compared, so it is found that they lead to satisfactory agreement for a wide energy interval in the continuum.