Research and development programs were supported in 12 inner-city, suburban, and rural schools in eastern Colorado. Success of the programs rested on the effects of conventional bonding and belief (social control theory), friendship choices (differential association theory), and opportunities and life chances (strain theory). Each school's program plan and implementation, methods of assessment, and results are discussed. Results indicate that school programs can reduce delinquent behavior and make students better citizens in and out of school, improvements in instruction and school climate do not guarantee better behavior by students, and a focused program has a better chance to succeed than one which tries for a multitude of school improvements at once. 2 figures, 6 tables, references.