THE SIMULATION IS RUN OVER DISCRETE 1-MONTH PERIODS INCORPORATING THE COURT, CORRECTIONS, AND LAW ENFORCEMENT SYSTEMS. THE MODEL IS USED TO DETERMINE THE EFFECT OF VARIOUS SENTENCING STRATEGIES INVOLVING THE CERTAINTY AND SEVERITY OF PUNISHMENT. THOSE SENTENCING POLICIES WHICH CORRESPOND TO THE GREATEST ESTIMATED CRIME CONTROL EFFECT ARE IDENTIFIED AS OPTIMAL SOLUTIONS. EXTENSIONS OF THE ANALYSIS ARE DEVELOPED TO SPECIFY THE INCAPACITATIVE AND DETERRENT EFFECTS EMBODIED IN PREVAILING AND OPTIMAL SOLUTIONS. IN ADDITION, RESULTS OF THE MODEL ARE PRESENTED FOR DATA BASES ORIGINATING IN DIFFERENT GEOGRAPHICAL AREAS, AND A DETAILED COMPARISON IS PERFORMED. METHODS FOR THE VALIDATION OF THE MODEL ARE PRESENTED AND VALIDATION RESULTS ARE OUTLINED. RESULTS INDICATE THAT LITTLE CAN BE DONE IN THE WAY OF IMPLEMENTING POLICY TO CONTROL CRIME. IN FORMULATING THIS MODEL OF THE CRIMINAL JUSTICE SYSTEM, A LARGE BODY OF LITERATURE IS INTEGRATED IN AN EFFORT TO EXTRACT THE ADVANTAGES OF EACH APPROACH AND MINIMIZE THE SHORTCOMINGS OF THE MODEL. BY COMBINING FORECASTING EFFICIENCY OF TIME SERIES ANALYSIS AND THE LOGICAL INTERRELATIONSHIPS OF EARLIER DEVELOPMENTS, THE FORMULATION HAS IMPARTED A DYNAMIC CHARACTER TO FORMULATIONS WHICH PREVIOUSLY COULD BE ANALYZED UNDER STATIC CONDITIONS ONLY. RESULTS STEMMING FROM THE MODEL SHOULD PROVIDE INSIGHTS INTO PREVAILING JUDICIAL POLICIES, AS WELL AS SUGGEST NEW DIRECTIONS FOR POLICY IMPROVEMENT. BASIC STEPS FOR EXECUTING THE MODEL ARE OUTLINED, AND APPLICATIONS OF THE MODEL ARE DOCUMENTED IN CASES WHERE THE ACTUAL DATA ARE EMPLOYED FROM THREE STATES. COMPARATIVE AND SENSITIVITY STUDIES ARE PROVIDED. RECOMMENDED EXTENSIONS OF THE MODEL WOULD EMBODY THE DERIVATION OF A CLOSED FORM TRANSIENT EXPRESSION FOR THE BASIC EQUATION. A FINAL EXTENSION WOULD INVOLVE THE REBUILDING OF THE FORECASTING MECHANISMS WITHIN THE MODEL WITH EACH POLICY CHANGE. STATISTICAL FORMULAS AND EQUATIONS COMPRISE THE BULK OF THE WORK. TABLES AND A BIBLIOGRAPHY ARE INCLUDED, WITH AN APPENDIX CONTAINING ADDITIONAL DATA ON PROGRAMMING THE MODEL. (AUTHOR ABSTRACT MODIFIED--LWM).