ObjectiveThis article offers an examination of the notion of truth in psychoanalysis, specifically in the work of S. Freud and J. Lacan. Far from being linked to notions of accuracy, certainty, or objectivity, as it the case in the field of science, subjective truth offers itself for psychoanalysis from its own perspectives. By giving itself the goal of elucidating the truth of the subject and not the disappearance of symptoms, psychoanalysis also deviates from any reference to reality and its standards, and defines a field and criteria that are unique. MethodThis reflection is based on a chronological and thematic reading of the texts of S. Freud and J. Lacan; it does not claim to be exhaustive but makes it possible to trace the outlines of an elaboration where points of continuity and rupture are marked between the authors, but also within in one's elaboration. ResultsFor Freud, subjective truth is connected with the deciphering of the unconscious. Symptoms, dreams, and slips of the tongue are understood by opening up to an unintentional discourse, which is that of free association. The truth arises only by surprise, only manifests itself in fragments, only lets us grasp it after the fact. The work of psychoanalytic treatment leads to a reconstruction of the subject's story based less on accuracy than on plausibility, which earns the subject's conviction. For the Lacan of the 1950s, taking part in a re-reading of psychoanalysis based on linguistics, the dimension of truth is consubstantial with the dimension of language and therefore of the treatment: the subject cannot avoid confronting it from the moment s/he speaks. The speaking truth states a truth about which the subject would prefer to remain ignorant. The truth of the subject is that of a repressed desire, expressed in the field of the Other in the form of an inverted message. Rereading Epimenides's Paradox sheds light on the fact that truth belongs to the field of enunciation, the subject's own field. DiscussionWith Lacan, a second, more complex conception of truth emerges from the second half of the 1960s, when the possibilities conferred on speech and on the Symbolic recede in favor of a conception that recognizes an ever-expanding role of the Real and of the impossible. It is a truth designated as lying-truth, which can only be “half-said,” that of a truth recognized as horror, with the discovery of the unnamable thing, the objet-a, a part of the (speaking) body that is at once the most foreign and the most intimate. ConclusionThe subject's truth is not in itself, but in an object of a veiled nature: the objet-a that emerges in analytic discourse as impossible to say. This object is both the cause and the product of speech. The analyst's role is not to take pleasure in the truth of the analysand, which the former brings up as a question, without taking a position on the answer, which the analysand provides. At the end of a cure, the analysand can hope to have produced knowledge about her/his jouissance, irremediably of the order of fiction, which comes as close as possible to the impossibility of bearing witness to the Real.