NIBRS data from 1995 to 1999 were used to identify the study population of elderly assault and homicide cases. Data on homicides and aggravated assaults were extracted. Data on victims, offenders, and offenses were merged by linking variables to create the database. Weights were created to avoid multiple victim counts. Variables collected included the incident date, location, and circumstances; age, sex, and race of both victims and offenders; relationship of victim to offender; and weapon used. To avoid sparseness of data when analyzing victims aged 65 and older, categories for some variables were combined to yield adequate sizes. Descriptive and crude analyses were used to detect differences between homicides and assaults for elderly victims and victims of all ages. Pearson chi-square test and associated p values were used to examine the association between variables. Analyses that compared homicides and assaults of elderly victims (aged 65 and older) to younger victims were stratified over different covariates to identify possible confounder and variable interactions. There were 4,778 homicides and 389,457 victims of aggravated assault reported in the NIBRS population from 1995 to 1999. Males, Whites, and 15- to 39-year-olds were the predominant victims of both homicide and assaults. Persons aged 65 and older composed a small percentage of victims, but there was a higher proportion of homicides in this group than among younger victims ages 15 to 39. The elderly had a higher risk of death from assault than younger age groups, and there was evidence of an excessively high risk of fatality in both felony-related and firearm-related assaults for younger and older victims. Assaults that involved family/known persons were also associated with a higher fatality risk than incidents that involved strangers. Primary prevention strategies should focus on decreasing exposure to assault, especially in felony-related situations. This should involve making elderly persons less appealing targets for criminal offenders. This could involve educating the elderly on being aware of their surroundings, avoiding confrontation and risky situations, and participating in self-help groups and block clubs. Mandatory reporting of the abuse of elderly persons promises to be an effective means of reducing domestic violence that involves elderly victims. 5 tables and 69 references