Trouble with teaching and learning Argyris's Models I & II? I learned, over several semesters, how to teach Argyris's Models I & II in MBA classes devoted to implementation of innovations and strategies. Readers familiar with Chris Argyris's work on defensive reasoning are familiar with Models I & II. Model I shows how one's learned predisposition toward defensive reasoning gets in the way of problem solving and contributes to mistrust and error. Model II serves as a useful guideline for thinking about overcoming defensive routines and committing to publicly testing inferences in ways that build trust, foster learning, and solve problems among potential collaborators. My early attempts to teach the models in class produced low levels of learning. Here, I report the strategy I developed to produce higher levels of learning among a broader segment of graduate students. The strategy will interest students interested in developing a deeper understanding of Argyris's work on Models I & II. My primary purpose, however, is to speak to the practical realities of instructors who either assign Model I & II -related material in their classes or used to do so and given up because of the low level of learning produced; and would appreciate knowing about a pedagogical approach I have successfully used in my classes. Much of the teaching strategy emerged from a concerted effort to knit together my musings on what I wish Argyris had said about this . . .