The hypotheses are as follows. (1) Homicide offender rates for the elderly will be relatively stable across jurisdictions, whereas the overall homicide offender rate will fluctuate sharply. Data on elderly rate versus nonelderly rate for three types of variables -demographic/economic factors, homicide offender rates for other age groups, homicide offender rates for males, females, blacks, and whites -- contradict this hypothesis and indicate that elderly offender rates closely associate with a State's poverty level. (2) Homicides by the elderly will be disproportionately 'domestic' in terms of victim-offender relationship. Data confirm this hypothesis. (3) Homicides by the elderly will be disproportionately intraracial. This appears to be true for elderly black offenders but not necessarily for whites. (4) Circumstance/motives differ sharply by sex and race, and differences by sex and race found for other age groups will also be found for the elderly. (5) Homicides by the elderly will be predominantly incidents involving a single victim and offender. (6) Elderly offenders will be more likely to kill those of their own age than younger offenders. (7) Elderly offenders will be more likely to use firearms than offenders in other age groups. (8) Monthly patterns of homicides by the elderly will be similar to those of other age groups. Overall, the data indicate that the elderly offender is affected but to a lesser extent, by the same sociological forces creating different rates of murder from State to State and that the incidence of homicide varies by sex and race. Overall, the most significant question about elderly versus nonelderly homicide offenders not why they commit significantly different kinds of homicides (they do not), but why they commit fewer homicides of the same kind. Twelve references are provided.