In this paper we perform a simulation procedure of testing models for journeys to work. The testing regime is carried out on a number of such models, mainly within the class of gravity models. We test the models on synthetic populations constructed from an aggregated set of a large number of worker subcategories, reflecting for instance different qualifications. Each subcategory is constructed from a gravity model where the population size and parameters are drawn from random distributions. The advantage of this approach is that a large number of tests can be carried out repeatedly to test the response of different kinds of models. We test how specific attributes of the spatial structure and worker heterogeneity are captured by different modelling alternatives. In addition we find that some model formulations falsely tend to report significant contributions to characteristics that were not taken into account in the data generating simulation process. This illustrates the imminent risk of drawing wrong conclusions in empirical work.