To compare predictive validity of the Wilson Sims Fall Risk Assessment Tool (WSFRAT) with that of the Morse Fall Scale (MFS) in psychogeriatric inpatients. Psychogeriatric patients from Shatin Hospital, Tai Po Hospital, Castle Peak Hospital, and United Christian Hospital who had fall incident between April 2019 and April 2020 were identified. Their fall risks were assessed by the WSFRAT and the MFS, and their falls incidents during hospitalisation were recorded. Patients were classified as having high fall risk when their MFS score was ≥45 and when their WSFRAT score was ≥7. Sensitivity, specificity, and positive and negative predictive values of the two scales were calculated. We identified 183 (90 male and 93 female) psychogeriatric patients aged ≥65 years who had fall incident and were assessed by both the WSFRAT and the MFS during the study period. Among the 183 patients, four sustained a fall during hospital stay, giving a prevalence of 2.19%. All four patients were classified as having high risk of fall by WSFRAT, but only two of them were classified so by MFS. The sensitivity of WSFRAT was 100%, which was higher than the 50% by MFS, but specificity of MFS was higher than that of WSFRAT (45.81% vs 54.75%). WSFRAT is a better fall risk assessment scale for psychiatric inpatients than MFS, because of higher sensitivity (100% vs 50%). It has items specific to psychiatric patients and should replace MFS in psychiatric settings.