The objective of this study was to determine whether participation in clinical trials affects long-term outcomes in Alzheimer's disease (AD). Participation in clinical trials for persons with dementia is often justified on the grounds that patients benefit from the medical oversight typical of trials, even when experimental agents do not demonstrate short-term benefits. This claim has not been rigorously assessed. Of 215 community-resident subjects enrolled in a prospective study of outcomes in AD, 101 participated in randomized clinical trials (RCTs) during the first 2 years of follow-up. These subjects were compared with subjects who met eligibility requirements for RCTs but did not participate (N = 57) and with subjects who were ineligible (N = 57), over a total of 3.5 years of follow-up. Survival analyses assessed risk of death, nursing home placement, and incident functional deficit end points, adjusting for baseline differences. Subjects who participated in RCTs were younger and more highly educated. Mortality, risk of hospitalization, number of medical examinations conducted by study physicians, and onset of severe functional deficit did not differ between the groups, but risk of nursing home admission was significantly lower among RCT participants compared with eligible nonparticipants and ineligible subjects (16.8% versus 36.8% and 31.6%, respectively [p = 0.01]). The difference in risk of nursing home placement may represent a long-term, drug-related benefit to patients, a selection effect (caregivers of patients who participate in RCTs differ from caregivers of patients who do not), or a positive effect on caregivers (greater contact with a medical service may be associated with better care-giving outcomes). Further research is required to assess these effects.