In recent years a Danish debate on the use of traffic models have taken place in the professional community. IFP has among others participated intensively in this debate. One of the conclusions has been that many reminiscences of the early development of traffic models still exist - despite the recent development in computer and software technology, as well as theoretical development. A number of fundamental problems are: 1) That the coherence between sub-models seldom equals the road users and passengers decision-making process. 2) That the use of variables in different sub-models seldom are consistent with each other. 3) That advantages and disadvantages with the sequential versus other more recent model approaches have not been discussed thoroughly. 4) That people do not act rational as most models assumes. 5) That supply models (e.g. matrix estimation, route choice and traffic assignment) are too simplified in many decision making context. In phase 2 of the GIS-T programme, the above problems are dealt with in more fundamental discussions, while the following sub-models are dealt with more thoroughly; 1) Route Choice Models, 2) Matrix Estimation Methods and 3) Probit models for mode choices.