This study sought to determine the relationship between behavioral disturbance and functional status in a longitudinally studied sample of patients with Alzheimer's disease (AD). One hundred and forty-nine patients meeting NINCDSยฑADRDA criteria for probable AD were followed for an average of 37.3 months, with follow-up assessments every 6 months. Subjects were seen at the Alzheimer's Disease Research Center clinics at the Mt Sinai Medical Center, New York, and the Veterans Aairs Medical Center, Bronx, New York. Measures included the Physical and Self-Maintenance Scale (PSMS) and Instrumental Activities of Daily Living Scale (IADLS) of Lawton and Brody and the cognitive and non-cognitive subscales of the Alzheimer's Disease Assessment Scale (ADAS). For each patient the assessment at which they had their most severe non-cognitive symptoms as measured by the non-cognitive part of the ADAS (ADAS-NC) was determined. ADAS-NC scores at that assessment were correlated with IADLS and PSMS scores at the same assessment and at the next assessment 6 months later. While there was some modest association of ADAS-NC scores with functional impairment using pairwise correlation coecients, none of the correlations remained signiยฎcant when the severity of cognitive impairment was controlled statistically. Findings were not signiยฎcantly changed when drug status was controlled. These results suggest that behavioral disturbance, while very troubling to caregivers and patients, does not substantially worsen functional ability beyond the contribution of cognitive impairment in AD. Together with previous results indicating that non-cognitive symptoms in AD are episodic and ยฏuctuating rather than progressive, the present data suggest that interventions for non-cognitive disturbances in AD should be viewed as ways to increase patient comfort, safety and ease of care and not as ways to improve functional autonomy. The latter can be achieved only by improving the progressive cognitive deยฎcits of AD.