In “Constructions in Analysis”, Freud addresses the issue of securing the veracity of individual constructions in the analytic setting. The concern, Freud notes, is that an analyst's construction in therapy is unassailable, for a patient's agreement is confirmation and a patient's disagreement is a signal of repression, which is also confirmation. This undertaking is an examination of Freud's attempts to overcome the problem of unassailability in “Constructions in Analysis” and in other works. The difficulty of the problem of knowing the correctness of a construction was such that Freud often appealed to analogies— specifically, archeological reconstruction and puzzle making. Were these analogies arguments or mere heuristics? Overall, I argue that Freud did not believe that any construction was unassailable, but that each was a hypothesis in need of confirmation and confirmation of every construction occurs ultimately, for Freud, only at the close of therapy. I end with an answer to the question: To what extent does Freud's psychoanalytic therapy at its completion validate its constructions?