Traditional theories of “adolescent risk taking” have not been validated against recent research indicating that youthful traffic crash, violent crime, felony crime, and firearms mortality rates reflect young people’s low-socio-economic status (SES) compared with older adults’, not young age. Aside from a small number of recent, conflicting studies, the literature gap on this key issue remains. The present study examines the 54,094 homicide deaths, including 41,123 gun homicides, victimizing California residents ages 15 to 69 during 1991 to 2012 by poverty status. Without controlling for poverty, homicide rates display the traditional age-curve peaking at 19, then declining. When poverty is controlled, the traditional age-curve persists only for high-poverty populations, in which young people are vastly over-represented, and homicide rates are elevated for all ages. This finding reiterates that “adolescent risk taking” may be an artifact of failing to control for age-divergent SES.