Witnessing violence can negatively affect students' outcomes, but their friendships may mitigate those effects. This study investigated the extent to which seeing someone shot in the last year related to seven school-related outcomes. Additionally, it examined the extent to which prosocial and deviant friends moderate these relationships. This study used data from three waves of the LONGSCAN study (N = 962), employing a series of student fixed effects models to estimate the relationships of interest. This study found no evidence that seeing someone shot in the last year was related to changes in school-related outcomes. Limited evidence of a moderating effect of friends was present, although in unexpected directions. The findings suggest that the effects of childhood trauma are complex and point to the need for a context-specific understanding of exposure to gun violence.