Social bonds and social organization theory are combined to examine schools' roles in regulating youth gang involvement in the context of community changes--an integrated approach to analyze psychological and sociological influences simultaneously. A subsample (N = 269) from Gottfredson et al.'s (A national study of delinquency prevention in school final report, Gottfredson Associates, Inc., Ellicott City, MD, 2000) national sample of schools to examine school-based, gang-prevention programs was combined with U.S. Census data to model the effects of community changes and school social organization on student's probability of gang involvement. Changes in the community's level of concentrated disadvantage and racial heterogeneity were related to a school's level of disorder. The probability of a student being gang-involved almost triples if the student attends a school one standard deviation below the sample mean School Social Bonds score; it doubles if the student attends a school one standard deviation above the sample School Disorder mean. Student characteristics partially mediated the relationship between School Disorder and gang involvement. Findings support extending social bonds theory to the school level. School-based, gang-prevention efforts may benefit from climates characterized by prosocial bonds and low social disorganization, especially for schools in highly-disadvantaged and/or demographically-changing communities.