The aim of this thesis is to verify the applicability of a discrete choice model for the calibration of a destination choice model based on home-bound trips to work. The challenge is to define an appropriate selection procedure to minimize the number of potential destinations and to identify those variables having a significant impact on the model result. To limit the number of possible alternatives, the following approaches were defined: (1) Random selection of a certain number of destinations (2) definition of a selection procedure considering distance, travel time or utility classes (3) definition of limit values, (4) consideration of dominance criteria. It is a matter of common knowledge, that people are not very flexible in their choice of workplace, however, a correlation between the place of residence and the workplace can be assumed, defined by variables describing the quality of transport connection (trip length, travel, time, etc.). The attractiveness of a destination is taken into account by the additive logarithmic number of jobs in the utility function of the discrete choice model, i.e. the probability of choosing a traffic cell increases with the number of jobs located there. The results show that no systematization of the selection procedure improves the model results significantly, i.e. a random selection for the calibration of the destination choice model seems to be sufficient, although the restriction to "plausible" alternatives could be appropriate (e.g. limitation to a maximum distance), especially for large transport networks. Using economic sectors to describe the attractiveness of a destination leads to a better understanding of the mobility behaviour of specific population groups. Suitable variables to describe the quality of the transport connection are all time-related variables, such as average travel time by car or public transport. A distinction of the riding time in the vehicle and the transfer times of public transport is appropriate.