Objective: When reminded of their neurological history, mild traumatic brain injured (TBI) students underperform on neuropsychological tests. To date, this “diagnosis threat” (DT) phenomenon has mainly been studied with a non-clinical and high-functioning population (university students). The aim of this study was twofold: to study this phenomenon with neurological patients and to examine the mechanisms responsible for underperformance. Method. Patients (18–55 years-old) who had sustained a TBI or a stroke were recruited from ambulatory and hospitalized cares, and then assigned to one of three conditions: Patients attention was drawn on (1) their neurological disease and the neuropsychological components of the upcoming tasks (DT group); (2) their intact sensory capacities and the sensorial components of the tasks (Neutral group); or (3) their better cognitive abilities compared to Alzheimer disease patients (Stereotype boost group). After these instructions, patients carried out cognitive tasks and completed questionnaires. Results: Preliminary analyses (n = 18) showed that, on the z-score of executive functioning, the DT group performed worse than both the neutral group (p = .03) and the stereotype boost group (p = .05), but did not differ for the attentional and memory scores. Instructions also had an impact on cognitive self-efficacy, with the neutral group demonstrating greater score than the negative one (p = .08). Furthermore, the self-efficacy score tended to correlate with the score of executive functioning (r = .37). Conclusion(s): Results show that the DT phenomenon has an impact on cognitive performances in clinical setting, at least on executive functions, which are usually demonstrated to be the most sensitive to stereotype effects.