The model also makes projections according to inmate classification levels and estimates the impacts of pending legislation and policy changes on offender populations. The Inmate Classification Projection Model is a probabilistic model based on Markov Chains. It groups the inmate population into existing inmates, new intakes and releases, and handles each group separately according to classification levels and parole violation rates. The model requires data on the numbers of offenders, numbers of defaultors with new commitments, the distribution of inmate classifications, projected intakes, and other factors. The computer program needed by the model is written in the SAS language. The output of the model is a 10 by 7 matrix that includes the 6 classification categories and the total population at the end of each projection year. List of data elements and copy of the computer program.