Source: (2004) Child Abuse & Neglect. 28: 267-287.The model of family group conferencing (FGC) for decision-making in child welfare has spread from its birth place in New Zealand to many other countries, including several in Europe. Some have contended that the wide popularity of FGC rests more on procedural and implementation data than on outcome evidence. This has led them to stress the need for research on long-term outcomes. In this context, Knut Sundell and Bo Vinnerljung report on results from a three-year follow-up of a trial FGC in ten local authorities in Sweden. Specifically, they compare outcomes for 97 children involved in 66 FGCs in 1996 and 1997 with results for 142 children from a random sample of traditional child protection investigations conducted by the Child Protective Services during the same period.