McGuire's inoculation theory employs an analogy to predict the effectiveness of defenses in inducing resistance to persuasion. In discussing this analogy, McGuire states that it is necessary to use beliefs which have not been attacked previously. Many studies have used inoculation theory to predict resistance with beliefs which have been attacked. We argue that McGuire is incorrect in his interpretation of the analogy and discuss an experimental test of different defenses' effectiveness over three belief levels.