Introduction: The Clinical Frailty Scale (CFS) has been validated internationally to predict adverse outcomes and mortality. Frailty assessments in the Emergency Departments (ED) are challenging to due to a lack of training and time. We studied the use of a tablet-based CFS that used graphics and short descriptors to assist choice of the 9 frailty categories. Methods: We conducted a prospective observational cohort study of people >65 years seen in the ED of 3 Canadian academic centers. We excluded critically ill patients, and those with significant visual impairment or inability to communicate in English or French. We compared agreement on the tablet-based CFS between 4 categories of assessors: Patients, ED Physicians, trained Research Assistance and Caregivers using the kappa statistic. Results: We enrolled 274/380 eligible patients who provided complete data (72.1%). Their average age was 75.8 years, and 48.9% were female. Their median MOCA score was 23/30 (IQR=17-26) and their median OARS was 26/28 (IQR 22-28). Agreement between physicians and research assistants was good (κ=0.60, 95% CI 0.50-0.70), as was physician-caregiver agreement and patient-caregiver agreement (κ=0.66, 95% CI 0.40-0.93). Agreement between physicians and patients was only moderate (κ=0.47, 95% CI 0.36-0.58). Conclusion: There was less agreement between physicians and patient self-assessments for the CFS compared to physicians-research assistant agreement and care-giver patient assessments of frailty. Future research should validate whether MD, patient, or caregiver rated CFS have higher predictive validity.