The Karnofsky Performance Scale (KPS) was evaluated in a geriatric outpatient population with regard to three issues: its strength of association with widely used and validated geriatric instruments; its ability to predict patient outcomes; and its ability to serve as an identifier of high-risk patients. The 134-patient sample was given a comprehensive geriatric assessment which included the KPS, the Activities of Daily Living (ADL) scale, the Instrumental Activities of Daily Living (IADL) scale, and other psychosocial and sensory tests. The KPS, ADL, and IADL were significantly correlated with each other, and the KPS showed the strongest associations with other functional measures. The KPS was also highly predictive of outcomes, performing better or equally well as the ADL and IADL. The KPS designation of high- and low-risk groups resulted in statistically significant score differences between groups in all but one assessment area, demonstrating better ability to discriminate than either the ADL or IADL. Thus, the KPS was shown to serve as an effective proxy score for a patient's health and functional status. It also was a significant predictor of hospitalizations, survival time, community residence, and institutionalization. Finally, the KPS was shown to adequately distinguish risk groups to aid in the targeting of services to ambulatory geriatric patients.