The preponderance of studies on frailty assessment in critically ill adults have used the Clinical Frailty Scale (CFS) to quantify frailty and previous research suggests that surrogates were more likely to be optimistic than physicians in their CFS scores. Whether discordance between surrogates and physicians was relevant to prognosis has been underexplored. Therefore, in a prospective observational cohort of 298 critically ill older adults, we aimed 1) to describe factors related to discordance and 2) to estimate the relationship between such discordance and hospital mortality and other short-term outcomes. Discordance between surrogates and physician was present in 89/298 (29.9%) and independently associated with a higher risk of hospital mortality. Discordance was not associated with markers of intensity of treatment such as intubation, blood transfusion, incident dialysis for acute renal failure and prolonged hospital length of stay. Understanding factors relevant to discordance between physicians and surrogates may lend further insights into short-term prognosis for older adults with critical illness.