In this study, I employ geospatial techniques to assess the impact of school attendance zone "gerrymandering" on the racial/ethnic segregation of schools, using a large national sample of 15,290 attendance zones in 663 districts. I estimate the effect of gerrymandering on school diversity and school district segregation by comparing the racial/ethnic characteristics of existing attendance zones to those of counterfactual zones expected in the absence of gerrymandering. Results indicate that the gerrymandering of attendance zones generally exacerbates segregation, although it has a weaker effect on the segregation of Whites from Blacks and Hispanics. Gerrymandering is particularly segregative in districts experiencing rapid racial/ethnic change. However, gerrymandering is associated with reductions in segregation in a substantial minority of districts, notably those under desegregation orders.