This paper addresses how terrain model requirements for research and development simulations differ and propose methods for comparing models implemented in dissimilar formats. Simulations for evaluating prototype mission equipment packages for aircraft often require high resolution digital elevation models (DEMs) and complex cultural feature models, primarily to support computer generated imagery. However, manned research simulations are now placing additional demands on the elevation and feature models. Terrain models are typically described in reference to their specific implementations that makes comparison of models difficult. Identifying terrain model requirements for a simulation involves comparing the cost of the implementation to the utility of the model. One level of modelling is not appropriate at all stages of development and resources can be better utilized if the appropriate models are selected at each stage. In addition to presenting methods for comparing models, this paper discusses how the modelling requirements change during the development of a complex simulation.