Dysphagia occurs in various diseases and constitutes a major concern in patients with psychiatric disorders. The Mann Assessment of Swallowing Ability (MASA) comprises 24 clinical parameters designed to identify swallowing disorders. One item in MASA, the "gag reflex", involves an unpleasant stimulus, which means that it is often omitted when the test is administered. The aims of this study were to determine the presence/absence of dysphagia in patients with psychiatric disorders using the MASA and determine its diagnostic accuracy when the gag reflex item was excluded in patients with psychiatric disorders. The study participants comprised patients admitted to a hospital psychiatric ward in whom dysphagia had been suspected based on oral intake status. The following items were determined: age, total MASA score (23 out of 24 items, giving a score out of 195 points), body mass index score, milligram equivalents of chlorpromazine, and the Food Intake Level Scale score. The patients were divided into two groups according to the presence or absence of swallowing problems as assessed by videoendoscopic or videofluoroscopic examination. The scores for each item investigated in the MASA, including the total score, were compared between the two groups. Receiver operating characteristic curve analysis was carried out to determine the optimum cut-off value. The total MASA score, which excluded the "gag reflex" item, was lower in the problematic swallowing group than in the non-problematic swallowing group. The MASA scores for cooperation, respiratory, dysphasia, tongue coordination, oral preparation, pharyngeal phase, and pharyngeal response tended to be lower in the problematic swallowing group. Furthermore, an optimum cut-off value of 169 points (sensitivity, 0.92; specificity, 0.68; likelihood ratio, 2.84) was identified. These results indicate that the cut-off MASA score is effective in screening for dysphagia, even when the "gag reflex" item is excluded.