The article first documents the increased use of incarceration in both the adult and juvenile systems under the law reforms of the last generation, with attention to the impact this has had on State budgets. The theoretical basis for assuming that these reforms should result in lower crime is examined, and empirical evidence relevant to this issue is analyzed. The research provides little support for the assumption that punitive reforms have reduced crime beyond reductions attributable to incapacitation. The article then introduces scientific research on adolescent development that underscores the impact of correctional interventions and settings on developmental trajectories and reoffending. This developmental research reinforces the conclusion that for most juveniles, long incarceration increases the social cost of crime and should be used only when public safety is at stake. The research also argues that some community programs can lower these costs. The concluding section of the article considers the implications of development research for the legal concepts of retribution and proportionality. The authors conclude that retribution must be based on culpability as measured by developmental research pertinent to behavioral controls. This perspective is especially important in the development of a system that responds appropriately to juveniles who commit crimes. 193 notes