Background Discussing end of life preferences can be beneficial, and it is thought that the best time to have these conversations is usually when people are well. This review aims to establish current evidence for the effectiveness of community-based interventions to encourage people to consider, and to discuss with those closest to them, their preferences for end of life care or what they wish to happen after their death. Methods A systematic literature review was undertaken. A systematic search was conducted using Scopus and Google, and academic experts were contacted. Studies were included if they evaluated interventions intended to encourage people to discuss their end of life preferences with those closest to them, or to address known barriers to these discussions. Reported outcomes had to relate to attitude or behaviour change in the target group, or target group perceptions of the intervention. Studies were excluded if the intervention targeted only people with a life-limiting illness, or intended specifically to facilitate communication of end of life preferences between patients and healthcare staff. Studies were systematically described and assessed for quality. There was no attempt to combine results of different studies. Results The Scopus search identified 5743 citations, and the Google search identified over 40,000, of which the first 40 pages were scanned. Five studies were included, four identified through the Scopus search and one from a book identified through Google. Three studies reported positive results, two were less positive. A peer education programme on end of life planning for older people, featuring small discussion workshops, was positively appraised by participants. An arts project bringing hospice users and school pupils together appeared to help normalise death for school pupils. A public information 'roadshow' engaged people using an informal questionnaire survey, facilitating conversations between people who participated together. Public lectures by physicians intending to promoting home death as a possibility were unsuccessful in changing attitudes at six months follow-up. A module on end of life planning delivered as part of 'expert patient' education programme on the management of chronic illness was not well received by participants. Conclusions Available evidence highlights the importance of actively engaging people rather than passively providing information, and of ensuring an appropriate context for interventions. However, data are limited and there is a need for more research and for sharing of best practice. [ABSTRACT FROM AUTHOR]