The sample consisted of 42 physically abused elders and 42 controls drawn from 3 model projects on elderly abuse funded by the Administration on Aging. The sites, based in Massachusetts, New York, and Rhode Island, offered casework services to elderly victims of maltreatment. Both groups were predominantly female, white, low-income, and Catholic. The abusers were much more likely than the controls to have mental and emotional problems and to abuse alcohol. No significant correlations between abuse and a history of family violence were found. Abused elders were neither more ill or functionally disabled than the control group, but were less impaired in several areas. However, the abusers were more likely to be dependent on their elderly victims than were the control relatives. While abusive families were more likely to experience stressors such as persons moving in or out of the house, these stressors were usually actions of the perpetrator directly related to the abuse situation. Other truly external events did not differ significantly between the two groups. The abuse sample had fewer overall social contacts and were less satisfied with those they had. Tables, diagrams, and 28 endnotes.