Abstract
This paper deals with some theoretical aspects of the method and a consideration of experiences. In part one the disadvantages of the sequential calculation of distribution and model choice, the estimation theory and some characteristics of the model are considered. In part two the experiences gained by usage of the method in several projects are discussed. This part deals with the form of the function in general, the influence of different assumptions about the costs of trips and the geographical transferability. Finally the consequences of the use of origin-destination matrices, which are only partly filled with surveyed observations, and the geographical transferability are mentioned.