Objectives: Caregivers of hospitalized older adults experience elevated levels of stress and are at risk of poor health outcomes. There is a lack of screening tools based on self-reported caregiver variables incorporating both protective and risk factors, for early identification of at-risk caregivers. This study reports the development of a caregiver-centric screening tool to identify risk of depression at admission and predicts 3-month risk of depression and quality of life amongst caregivers of older adults with an unplanned admission. Design, Setting and Participants: This prospective cohort study was conducted in the medical wards of a tertiary-care hospital from July 2015 to May 2017. We recruited family caregivers of patients aged 65 years and above who fulfilled the following criteria: a) unplanned admission, b) not residing in nursing homes; and c) requiring assistance in activities of daily living. Measurements: We examined 11 candidate caregiver variables (mastery, burden and nine demographic variables). Risk of depression (score ≥8 on Hospital Anxiety and Depression Scale (HADS-D) depression subscale) was the primary outcome, and was assessed during the index admission. Logistic regression models were used to identify risk factors and risk scores (weights). The total risk scores were then stratified into three risk levels. Predictive validity of the screening tool was assessed using 3-months post-discharge risk of depression and health-related quality of life (HRQoL). Results: The study included 274 caregiver-patient dyads. The mean (SD) age of the caregivers was 59 (10) years with 33.6% caregivers screening positive for risk of depression. The final model comprised three caregiver variables: mastery, burden and education. The total risk scores ranged from 0 to 6 and showed good discrimination (AUC:0.82, 95% CI: 0.77 to 0.87). Caregivers were classified into low-risk (0–1 points), intermediate-risk (2–4 points), and high-risk (5–6 points) groups, with corresponding rates of risk of depression (HADS-D≥8) of 10.7%, 44.6% and 73.3%, during admission. Relative risk rates of the intermediate- and high-risk group using the low-risk group as reference were 4.16 and 6.84 respectively. At 3-months post-discharge, the rates of caregivers at risk of depression or having poor HRQoL also increased corresponding to the three risk levels as per baseline, supporting the predictive validity of the tool. Conclusions/Implications: The caregiver-centric tool is a novel, practical, self-administered, relatively brief caregiver-centric instrument that can be used for rapid screening and stratification of caregivers at risk of depression. Uniquely, the tool comprised of assessment of protective factor (mastery) in addition to risk factors to provide a holistic assessment of the caregiver. It can be incorporated as part of older adults’ admission evaluation so that prompt intervention can be rendered to their at-risk caregivers.