AIM: To review and meta-analyse Mellow Parenting interventions for parent-child dyads at high risk of adverse developmental outcomes.

METHOD: Using Preferred Reporting Items of Systematic reviews and Meta-Analyses (PRISMA) guidelines, we extracted all published evaluations of Mellow Parenting and Mellow Babies programmes. We identified published studies with randomized controlled trials, quasi-experimental or within-subject pre-post designs. We incorporated 'grey literature' for unpublished publicly available evaluations. Effect sizes were calculated for impact of Mellow Parenting on parental mental health and child behaviour. Data were extracted on demographics, age of participants, country, and potential sources of bias.

RESULTS: We identified eight papers, representing nine data sets, from five of which we calculated effect sizes. There was evidence of a medium treatment effect of Mellow Parenting compared with comparison groups on maternal well-being and child problems. Drop-out from treatment was variable. However, data were heterogeneous and there was evidence of methodological bias.

INTERPRETATION: Our data give some support to claims for effectiveness of Mellow Parenting as a group intervention for families with multiple indices of developmental adversity. Given the methodological weaknesses of literature in the area, novel approaches are needed in future trials of low-budget complex interventions in non-commercial settings.