Data on subject selection biases in research with older adults are limited. We conducted two sample bias studies that assessed primary caregivers for dementia patients. The first study compared 46 caregivers, who required home interviews, with 63 caregivers willing to travel to the university hospital for assessment; all subjects had been offered free taxi service. Caregivers interviewed at home were helping more hours per day and were more depressed than those assessed at the university; in addition, the patients of caregivers assessed at home were more impaired and had greater personality changes. The second study compared available information from respite care users who chose to participate in our caregiver research project with those who chose not to participate. Nonparticipants provided care for more impaired patients. No demographic differences were found between groups in either study, and there were no differences in the chronicity of caregiving. The implications of these data for aging research are discussed.