INTRODUCTION: Frailty is thought to be associated with an increased risk of adverse health outcomes such as death and falls, but comparatively little is known about the impact of frailty transitions on the adverse health outcomes. Moreover, owing to insufficient sample size or a single-center study design, previous studies have not been sufficiently representative of elderly inpatients in China. This study aimed to provide estimates at the population level of the association between frailty transitions and adverse outcomes among elderly inpatients following discharge. METHODS: This was a large-scale multicenter cohort study conducted from October 2018 to February 2021. The FRAIL scale was used to estimate frailty status. Frailty transitions were derived by considering frailty status at baseline and the 3-month follow-up, which encompassed five patterns: persistent non-frailty, persistent pre-frailty, persistent frailty, improvement in frailty, and worsening of frailty. The outcome variables included mortality, falls, hospital readmissions, and Health-Related Quality of Life (HRQoL). Cox proportional hazard regression, generalized linear models and linear regression was used to examine the association between frailty transitions and adverse health outcomes. RESULTS: A total of 8,256 patients were included in the study, 40.70% of study participants were non-frail, 43.04% were pre-frail, and 16.27% were frail. Compared with patients who persistently non-frail patients, those who frailty improvement, persistent pre-frailty, worsening frailty, and persistent frailty showcased escalated risks of mortality within 2 years after enrollment [HR (95% CI): 1.32 (1.06-1.64)], 1.71 (1.37-2.13), 2.43 (1.95-3.02), and 2.44 (1.81-3.29), respectively. These groups also faced elevated hazards of 2-year falls [OR(95% CI): 1.586(1.13-2.23), 2.21(1.55-3.15), 1.94(1.33-2.82), 2.71(1.59-4.62)] and re-hospitalization risk within 2 years[OR(95% CI): 1.33(1.13-1.56), 1.56(1.32-1.86), 1.53(1.28-1.83), 2.29(1.74-3.01). The number of falls increased by 0.76 over 2 years in frailty-worsened patients and 0.81 in persistently pre-frail patients. The total days of rehospitalization increased by 0.35 over 2 years in frailty-improved patients, by 0.61 in frailty-worsened patients, by 0.66 in elderly in persistently pre-frail patients and by 0.80 in persistently frail patients. Moreover, patients exhibiting frailty-improved [-1.23 (95% CI: -2.12 to -0.35)], persistently pre-frail[-4.95 (95% CI: -5.96 to -3.94)], frailty-worsened [-3.67 (95% CI: -4.71 to -2.62)], and persistently frail [-9.76 (95% CI: -11.60 to -7.93)] displayed inverse correlations with the regression coefficients of HRQoL. DISCUSSION: Frailty-improved, worsened, persistently pre-frail, and frail inpatients face higher risks of mortality, falls, rehospitalization, reduced HRQoL than consistently non-frail inpatients. Screening for frailty among elderly inpatients can identify individuals at increased risk of adverse health outcomes.