An action training program that teaches inspirational communication of a vision as part of a training of charismatic leadership for managers is presented (1 1/2 days) and evaluated in 2 studies (N = 25 and N = 22). We used the research design "nonequivalent dependent variable design" (Cook & Campbell, 1979, p. 118) or "internal referencing strategy" (Haccoun & Hamtieux, 1994), which compared the trained behaviors (charismatic inspirational communication) with behaviors that were not trained (public speech) to control for testing and Hawthorne effects. The training had specific positive effects on those behaviors that were trained but not on those variables that were not trained. Good to excellent effect sizes appeared as a result of the training. We suggest that this research design is useful for evaluation of training effects within the constraints of commercial settings and, moreover, we argue that this design is in many ways superior to a nonequivalent nontraining control group design because it controls for testing effects and for effects that otherwise would need a pseudotraining control group.