An overall description of the planning process shows the joint involvement of the experimental and the analytical evaluations in systematic attempts to make rational choices for the future; the emphasis is both on making choices and on the orientation toward the future. The program chosen must be implemented on a pilot basis which enables an experimental postevaluation of its impact modeled on the classic random, controlled experiment. A particular model for analytical evaluation in the criminal justice system, the Justice System Interactive Model (JUSSIM), uses a computer in an interactive mode for the downstream flow of cases. This study describes in detail how JUSSIM operates, its structure, and a typical run. An extension model, JUSSIM 2, is intended for augmentation with recidivist feedback. Several states (e.g., California, Alaska) and cities (e.g., Pittsburgh, Philadelphia, and Denver), as well as some foreign countries (e.g., Canada and Sweden) have organized collection efforts to describe their systems as base cases for the JUSSIM model, and they intend to use a version of JUSSIM for their own planing. The JUSSIM model can also be used as a teaching tool in systems analysis courses; to show the distribution of costs (workload) associated with the various portions of the criminal justice system; to evaluate a wide variety of potential system changes; and to conduct cross-jurisdictional comparisons. The results of such evaluations should not be treated as conclusive. Possible future extensions of JUSSIM analytical models could include estimating the incapacitative effects of imprisonment, the mean duration of individual criminal careers, and the effect of deterrence associated with a particular sanctions policy. An appendix provides references on JUSSIM. Twenty-six references are appended.