Previous evidence on the impact of Family Group Conferencing in the UK cannot establish causality. This article presents results from the first UK randomised controlled trial of Family Group Conferences. All families entering pre-proceedings between September 2020 and May 2022 in twenty-one local authorities in England were eligible for inclusion in the trial. In total, 1,511 families were randomised and child outcomes were collected from administrative data until up to twenty-one months post-referral. Half of the families were randomly allocated to be referred for a Family Group Conference, on top of usual practice during pre-proceedings. The (pre-registered) primary outcome is the child’s looked-after status twelve months after the pre-proceedings letter. This was 8.6 percentage points lower among referred families (36.2 percent vs 44.8 percent). Children were also less likely to have care proceedings issued (59.2 percent vs 71.7 percent) and spent less time in care (87 vs. 115 days). However, there was no significant difference in whether a child’s living arrangement remained the same or changed in the months after the decision. These findings have implications for the support given to families during pre-proceedings; they provide evidence in favour of Family Group Conferences as a way to reduce the chances of children going into care.