This Note describes a first-generation experimental computer model, Scenario Agent, as developed in RAND's Strategic Assessment Center in early 1981. Section II describes the Strategic Assessment Center and the role the Scenario Agent plays in it. Section III describes the conceptual approach taken in developing the Scenario Agent. Section IV discusses the Scenario Agent's inner workings in some detail so both programmers and nonprogrammers can understand them. Section V illustrates the results of using the Scenario Agent in a demonstration of the Strategic Assessment Center and discusses ideas for future refinement, some of which are now being implemented. Four appendixes list names of scenario entities, give examples of a tabular scenario (tableau) and of a narrative scenario, and discuss rule-based modeling.