Previous research has found reduced mortality from aggravated assaults, attributed to medical care improvements. However, aggravated assault has limitations as a longitudinal measure of injuries from violence. The National Crime Victimization Survey (NCVS) can address this by providing estimates of serious injuries from criminal victimization. Their lethality trend is not compatible with the previous finding across 1973 through 1999, remaining stable rather than falling. After 1999, both Uniform Crime Reports (UCR)-and NCVS-based measures indicate increases in lethality. The trend differences raise serious problems of data choice for the researcher.