I constructed a unique set of data from over 300 California law enforcement agencies, in conjunction with large-scale education microdata covering the high school outcomes of over 3.8 million California ninth-graders from the classes of 2003 to 2014 to examine the extent to which estimated effects of violence exposure, coupled with significant differences in violence exposure rates, contribute to population-level differences in educational attainment. I find evidence for two important processes linking violence exposure and educational attainment. [1] High school dropout rates significantly respond to gun violence only if exposure exceeds a certain threshold. This threshold rule, coupled with differential exposure rates across student subgroups, leads to significant exposure effects on the dropout rates of black and Hispanic students, and no significant effects for white and Asian students. [2] Learning loss does not appear to be the primary mediator of exposure effects on dropout rates. This suggests that dropouts in high violence areas often have the academic capacity for educational attainment beyond realized levels.