A decision theoretic framework for the analysis of police behavior is first considered. In discussing optimal control of the police, it is advised that one of the basic premises of the decision problem is that agents possess a model of the system which formally describes the way the state of the world and the decisions of agents affect system outcomes. It is assumed that the functions of a police officer can be described by a system of discrete time, linear difference equations, which are presented. Also discussed are objective functions which represent the desirability of the outputs or their characteristics in terms of their weighted deviations from target values. The study also briefly considers some of the possible extensions of the basic control problem which account for uncertainty about the structure of the model. Data collection for optimal control applications to police behavior is then examined. Overall, it is concluded that a large scale effort to apply optimal control theory to police behavior should not be tried, while small scale studies using simplification strategies are recommended. It is suggested that studies focus on some relatively narrow dimensions of police behavior and apply optimal control theory at a microlevel. Twenty-seven footnotes are provided.