A simulation model is proposed to represent the channel of the Global Positioning System (GPS). Initially, a digital elevation model, building databases, and a vegetation model are processed to generate azimuth-elevation maps of path states (clear, shadowed, and blocked) for a large number of observers. At each simulation step, satellite positions are updated, and azimuths and elevations of paths from observers to satellites are calculated. Signal strengths and range errors are assigned to paths with the aid of random number generators for each path state. This information is processed to determine the statistics of the channel. It will be shown that model predictions are able to display good agreement with the results from an experimental campaign carried out for validation purposes. The simulation model will then be applied to a large number of observers deployed along two routes in densely urbanized areas in the city of Rio de Janeiro, Brazil (22.8° S, 43.3° W), with buildings displaying different height distributions to quantitatively show how the most probable number of available satellites, the probability that four or more satellites are simultaneously available (that is, that positions can be fixed), and the position errors change with the average building height and for vehicles in static and kinematic modes. In combination with the comparison between measurement and prediction results, this indicates that the simulation model is an efficient and flexible tool for studying and planning satellite-based location and navigation applications with accuracy and sensitivity, which will be used in future developments of mitigation techniques of multipath effects, leading to improved performance of intelligent transportation systems.