The traditional method of modelling shippers’ port utility behaviour has been the Random Utility models. The robustness of such models and their variants in producing unbiased estimates cannot be guaranteed in the presence of simultaneity. In this paper, we demonstrate how interrelationship among variables that describe shippers’ port utilization model can lead to simultaneity problem. Using port level shipment data from the Nigeria Ports Authority, we then show how this problem can be eliminated using Two-Stage Least Squares technique. Our post estimation test results confirm the robustness of this technique.