Historical and current child protection practice in Australia has been subject to severe criticism, yet in spite of a persuasive case for an alternative family inclusive form of child protection practice – something that offers the potential for improved client outcomes and improved worker job satisfaction – the model is not yet in widespread use. An international review of promising innovations in child protection, including examples of programmes from Australia, resulted in a list of eight identified trends. Common to all of these trends was evidence that good-quality evaluation had contributed to their recognition. If family inclusive practice is to gain greater acceptance, especially by bureaucrats, policy makers and holders of the purse strings, sophisticated forms of programme evaluation will be required. Such evaluations might emphasise practice-based research where researchers and frontline practitioners work together on all aspects of evaluation, including the initial design stage. While gold-standard randomised controlled trials may be included, methodological pluralism should allow inclusion of alternative approaches, such as realist evaluation and the involvement of practice research networks. The use of external evaluators might be usefully replaced with greater reliance on evaluation partnerships between evaluation experts (researchers) and frontline agency staff. Follow-up systematic reviews and meta-analyses might then allow the development of evidence-based arguments for change. Some Australian programmes have shown how rigorous evaluation practices have underpinned success and this evaluation focus could be emulated.