A detailed clinical example is used to illustrate how reality testing can create rather than foreclose opportunities for analytic investigation. It is proposed that effective analysis of transference within the treatment relationship requires close and explicit attention to considerations of reality. The author reconsiders certain conceptions of a special psychoanalytic reality, of regression in clinical analysis, and of the nature of free association, suggesting that they tend to discourage the realism necessary to productive psychoanalytic work. He underlines the importance of ongoing reference to therapeutic outcome as an aspect of reality.