Participatory research methods connect researchers with relevant communities to achieve shared goals. These methods can deliver results that are relevant to people's lives and thus likely to have a positive impact. In the context of a large and growing body of autism research, with continued poor implementation, and some evidence of community dissatisfaction, there is a powerful case for participatory autism research. In order to develop a framework for such collaborative working, a UK seminar series was organised and co-produced by autistic and non-autistic people with academic, practitioner and lived expertise. This article reports on the outcomes from the series, identifying five topics relevant to building a community of practice in participatory research: Respect, Authenticity, Assumptions, Infrastructure and Empathy. Each topic is connected to a specific example from within and beyond research, to inspire new practices in the field. We call for the development of participatory research skills among the autism research community and the facilitation of greater autistic leadership of, and partnership in, research. Such work, if delivered to a high standard, is likely to lead to better translation into practice and improved outcomes for autistic people and those who support them.