New research from New Zealand follows up the file outcomes over three years for 1003 young people aged 16 years who had family group conferences in 1998. Five hundred and twenty of them were interviewed. Observational data and interviews were collected from another 115 cases in 2001/2002. Findings are presented on the extent to which restorative goals have been implemented. Critical factors predicting outcomes are identified and the implications of these for policy and practice are discussed.The research demonstrates that the nature of the youth justice does affect critical outcomes for young people: both in terms of reducing offending and increasing the probability of other positive life outcomes. Restorative practices that include empowerment, the repair of harm and reintegrative outcomes make a positive difference while the extent of embeddedness in the criminal justice system, severe and retributive outcomes and stigmatic shaming have negative effects. There are also important findings for crime prevention that suggest the need to focus on support for families, the importance of educational qualifications and the need to respond effectively when children first come to the attention of the welfare and youth justice systems. Proposals are made for standards against which practice can be assessed. Author's Abstract.