Shifting criminal justice practices away from sligmalization and toward rcintegration is no small challenge. The innovation of community conferences in New Zealand and Australia has two structural features that are conducive to reinlegralive shaming: (a) selection of the people who respect and care most about the offender as conference participants (conducing to rcintegration) ; and (b) confrontation with victims (conducing to shaming). Observation of some failures and successes of these conferences in reintegrating both offenders and victims is used to hypothesize 14 conditions of successful reinlegration ceremonies. The spectre of failure haunts modern criminology and penology. Deep down many feel what some say openly—that 'nothing works': that despite decades of study and debate, we are no nearer deterrence than we ever were and/or that more 'humane' forms of treatment are mere masquerades concealing a descent into Kafaesque bureaucracy where offenders suffer a slow and silent suffocation of the soul. Worse still, we fear that even when something does work, it is seen to do so only in the eyes of certain professionals, while 'outside' the system ordinary citizens are left without a role or voice in the criminal justice process. This paper takes a different view. Rejecting the pessimism that pervades discussions about crime and punishment, it offers an optimistic view of at least one area—the punishment of juvenile offenders. It argues that it is possible to develop practices that 'work'—both in the sense of reducing recidivism and reintegrating offenders into a wider web of community ties and support and, at the same time, in giving victims a 'voice' in a fashion that is both satisfying and also socially productive. Further, it links a theory (reintegrative shaming) and a practice (the reintegration ceremony) which explain how to understand and how to implement this success. While there are elements that are quite distinctive about both the theory and practice of reintegrative shaming, there is also a great deal in common with the theory and practice of 'making amends' (Wright 1982); restorative justice (Cragg 1992; Galaway and Hudson 1990; Zehr 1990); reconciliation (Dignan 1992; Marshall 1985; Umbreit 1985); peacemaking (Pepinsky and Quinney 1991); redress (de Haan 1990) and feminist abolitionism (Meima 1990). We differ from abolitionists, however, in believing that it is right to shame certain kinds of conduct as criminal in certain contexts. The rest of the paper has two sections. The second section outlines some fieldwork