Just as formerly recalled screen memories may be remembered again in the course of psychoanalysis, so new screen phenomena can arise during the treatment process. This paper attempts to relate the process of reconstruction and the occurrence of a type of screening. It is proposed that, under certain circumstances, psychoanalysis can result in a mutually determined screen construction that both patient and analyst consider convincing and valid but which may, for the most part, actually be untrue as a result of the screen function. Screen memories arising during the reconstructive process resemble the déjà vu and déjà raconté situation in that the experience is felt with certainty to have actually occurred previously, although a degree of doubt about its having happened can consciously or unconsciously coincide. Factors contributing to the formation of screen constructions include regressive aspects of the analytic situation; superego elements and the experience of an injunction to remember; the wish to witness what took place in the past; direct or indirect suggestive influence by the analyst; a defensive identification with the analyst; and the analyst's theoretical orientation to reconstruction. Like a screen memory having some veridical content yet serving a masking function, a screen construction can have elements of truth. At the same time such constructions may function as a strong resistance of compromise against intense erotic, sadomasochistic, or narcissistic themes in the transference/counter-transference that were painful or conflictual in the past. If screen constructions are assumed to be historically real or valid, they may not be analyzed for their transferential screening role. A previously published case involving a presumably repressed memory of sexual molestation is considered from the perspective of the possible development of a screen reconstruction. Clinical and scientific aspects of seeking extra-analytic confirmation or falsification of reconstructions are discussed.