Functional homeostasis is the ability of an individual to withstand illness without loss of function. We investigate whether the level of functional homeostasis predicts adverse outcomes in the 6 months posthospital discharge in older men and women. A prospective cohort study was conducted in an acute care geriatric inpatient unit of a university hospital. Subjects included a consecutive series of patients admitted to the unit. The Functional Independence Measure (FIM) instrument was used to assess patients at four time points: preillness, hospital admission, hospital discharge, and 6 months postdischarge. Of the 122 subjects available for analysis, 64 (52%) experienced a decline in functional level from preillness to hospital discharge and were defined as having poor functional homeostasis, whereas 58 (48%) experienced no change or an increase in functional status and were defined as having good functional homeostasis. Those with poor functional homeostasis had a higher 6-month readmission rate to the hospital (59.4 v 39.7%; P=0.03) and a higher rate of any adverse outcome (78.1 v 50%; P=0.001) than those with good functional homeostasis. In logistic regressive analyses, functional homeostasis remained a significant and powerful predictor of adverse outcomes independent of actual level of function at discharge, age, gender, living status, and other factors that might influence outcomes. Change in functional status associated with an acute illness is an independent predictor of adverse outcomes and, in this study, a better predictor than actual level of function at discharge. Functional homeostasis is one approach to the quantification of the important but elusive concept of frailty in the elderly.