In the light of the declaration “Nothing about us without us” (Charlton, 2000), interest in co-production, and coproduced research is expanding. Good work has been done establishing principles for co-production (Hickey et al., 2018) and for good quality involvement (Involve, 2013; 4Pi, 2015) and describing how this works in practice in mental health research (Gillard et al., 2012a,b, 2013). In the published literature, co-production has worked well in qualitative research projects in which there is often methodological flexibility. However, to change treatment guidelines in the UK, e.g., the National Institute for Health and Care Excellence guidelines, and influence service commissioning, high quality quantitative research is also needed. This type of research is characterized by formal methodological rules, which pose challenges for the scope of co-production. In this paper we describe the significant challenges and solutions we adopted to design and deliver a coproduced randomized controlled trial of mental health peer support. Given the methodological rigidity of a randomized controlled trial, establishing clearly which methodological and practical decisions and processes can be coproduced, by whom, and how, has been vital to our ongoing co-production as the project has progressed and the team has expanded. Creating and maintaining space for the supported dialogue, reflection, and culture that co-production requires has been vital. This paper aims to make our learning accessible to a wide audience of people developing co-production of knowledge in this field.