Background: Patient/carer involvement in palliative care research has been reported as complex, difficult and less advanced compared to other areas of health and social care research. There is seemingly limited evidence on impact and effectiveness. Aim: To examine the evidence regarding patient/carer involvement in palliative care research and identify the facilitators, barriers, impacts and gaps in the evidence base. Design: Qualitative evidence synthesis using an integrative review approach and thematic analysis. Data sources: Electronic databases were searched up to March 2018. Additional methods included searching websites and ongoing/unpublished studies, author searching and contacting experts. Eligibility criteria were based on the SPICE (Setting, Perspective, Intervention, Comparison, Evaluation) framework. Two quality assessments on methodology and involvement were undertaken. Results: A total of 93 records were included. Eight main themes were identified, mainly concerning facilitators and barriers to effective patient and carer involvement in palliative care research: definitions/roles, values/principles, organisations/culture, training/support, networking/groups, perspectives/diversity, relationships/communication and emotions/impact. Evidence on the impact of involvement was limited, but when carried out effectively, involvement brought positive benefits for all concerned, improving the relevance and quality of research. Evidence gaps were found in non-cancer populations and collaborative/user-led involvement. Conclusion: Evidence identified suggests that involvement in palliative care research is challenging, but not dissimilar to that elsewhere. The facilitators and barriers identified relate mainly to the conduct of researchers at an individual level; in particular, there exists a reluctance among professionals to undertake involvement, and myths still perpetuate that patients/carers do not want to be involved. A developed infrastructure, more involvement-friendly organisational cultures and a strengthening of the evidence base would also be beneficial.