Centrifuge performance maps are useful for characterizing the separation parameters of a particular centrifuge over a range of operating parameters such as feed rate and cut. Typically, to develop a performance map, the internal hydrodynamics and isotope diffusion must be modeled, requiring knowledge of the centrifuge geometry and operating conditions. Due to the sensitive nature of centrifuge information, these parameters are usually not available. In some cases, some information about the performance of the centrifuge may be inferred from publicly available information. In these cases, a method for estimating the centrifuge performance over a range of operating conditions can be useful. In this paper, a semi-empirical method for developing a centrifuge performance map from known performance data is presented. The method is based on analyzing the centrifuge as a square cascade with three characterizing parameters. To verify the model, centrifuge separative power and separation factor are compared with results from the Pancake code over a range of feed rates and cuts. A cascade level comparison is used to further compare the two models. The comparison shows that the semi-empirical model is in good agreement with the Pancake code for most cases.