Evaluators of argument are frequently confronted by conflicting claims. While these claims are usually based on probabilities, they are often resolved with the accepted claim treated as though it were "true," while the rejected claim is treated as though it were "false." Scenario testing is the label applied to a set of procedures by which competing claims are evaluated as "probabilistic" scenarios. Resolution of a proposition by scenario testing requires that an accounting of each substantiated scenario be made. Scenario testing begins with the assumption that conflicting claims should not be resolved by disjunctive choices. Rather, it assumes that descriptions and predictions provided by the debaters in a round are, by their nature, probabilistic. As a consequence, evaluation of a policy, whether the single policy offered by the Affirmative or the combined policies offered by the Affirmative and Negative are most reasonably evaluated when the probabilities of their descriptions and predictions are taken into account. An optimum policy choice, therefore, represents the efficacy of a policy (or policies) across the range of scenarios it must address. This procedure equalizes the burdens of the advocate and counter-advocate and provides a set of guidelines for evaluating their conflicting claims. (Author/HOD)