This is the author accepted manuscript. The final version is available from OUP via the DOI in this record.In their article ‘A marriage made in hell: Child protection meets early intervention’, Featherstone et al. (2014) question the value of early intervention in preventing or addressing early signs of child maltreatment. In this article, we summarise and critique their main contentions. Among the issues we cover are the difference between intervention and support, the tension between fidelity and flexibility, the relative value of randomised controlled trials, the evidence of ‘what works’, the use of neuroscience, the place of innovation and the role of wider socio-economic factors. We are sympathetic to many of the points raised by Featherstone et al. but argue that they misrepresent early intervention, provide insufficient empirical support for their case and ignore evidence that runs counter to their views. We outline an alternative vision for child protection that addresses many of the concerns expressed while incorporating high-quality evidence on early intervention.Vashti Berry's time is supported by the National Institute for Health Research (NIHR) Collaboration for Leadership in Applied Health Research and Care South West Peninsula.