Abstract Purpose To examine the prognostic accuracy of the Clinical Frailty Scale (CFS) and Braden Scale (BS) separately and combined for 90-day mortality. Furthermore, to examine the effect of frailty on mortality depending on different levels of the Braden score. Methods The study included acutely admitted medical patients ≥ 65 years. We used an optimum cutoff for CSF and BS at ≥ 4 and ≤ 19, respectively. CFS categorized frailty as Non-frail (< 4), Frail (4–5), and Severely frail (> 5). Prognostic accuracy was estimated by the area under the receiver operating characteristic curves (AUROC) with 95% confidence intervals (CI). Cox regression analysis was used to compute the adjusted hazard ratio (aHR) for mortality. Results The mean age among 901 patients (54% female) was 79 years. The AUROC for CFS and BS was 0.65 (CI95% 0.60–0.71) and 0.71 (CI95% 0.66–0.76), respectively. aHR for mortality of CFS ≥ 4, BS ≤ 19, and combined were 2.3 (CI95% 1.2–4.2), 1.9 (CI95% 1.3–2.9), and 1.9 (CI95% 1.3–2.8), respectively. For BS > 19, the aHR for mortality was 2.2 (CI95% 1.0–4.8) and 3.5 (CI95% 1.4–8.6) for ‘frail’ and ‘severely frail’, respectively. aHR for BS ≤ 19 was 1.1 (CI95% 0.4–3.2) and 1.3 (CI95% 0.5–3.7) for ‘frail’ and ‘severely frail’, respectively. Conclusion Although CFS and BS were associated with 90-day mortality among older acutely admitted medical patients, the prognostic accuracy was poor-to-moderate, and the combination of CFS and BS did not improve the prognostic accuracy. However, the hazard of mortality across different levels of frailty groups were particularly increased among patients with high BS scores. Graphical abstract