Capital asset management is a critical component of the operation of transit systems. In particular, much interest has been generated lately regarding the development of rolling stock deterioration models that can predict the future condition of a fleet from the corresponding deterioration curves. Based on a rolling stock inspection data set from Athens, Greece, this paper presents the development of both an ordered probit model and a predictive discriminant function that can be invaluable tools in predicting rolling stock deterioration. This combination of models provides a way in which we can get both aggregate (system level) projections on future bus conditions and disaggregate (individual bus level) projections. Both of the methodologies used recognize the ordinal nature of condition ratings and link deterioration to a set of relevant explanatory variables such as bus age, mileage, and size. The results can be easily used in a number of practical situations, such as capital asset life-cycle cost analysis, optimal timing for bus replacement, and examination of the effect of different operational strategies on bus deterioration.