Abstract : The purpose of this study was to test whether tactical intelligence analysis could be improved by using an interactive Bayesian inference aid. The study was performed with experienced tactical intelligence analysts (officers and enlisted personnel) and inexperienced but trained analysts (officers and enlisted personnel) and inexperienced but trained analysts (officers who had completed their training at the Army Intelligence Center and School). Two hypotheses guided the study. The first hypothesis was that experienced and inexperienced analysts working with the aid would be better able to discriminate between the most and least likely opposing force courses of action than analysts working without the aid. The second hypothesis was that this improved performance would be the result of significantly less conservatism with than without the interactive Bayesian aid. Both hypotheses were confirmed. Discussions with experienced tactical intelligence analysts indicated that such a finding could have significant implications for improving tactical intelligence collection and analysis.