Background: Previous trials of interventions to support stroke survivors and their families in the community have had contradictory and inconclusive results. Using the MRC Framework for Complex Interventions we developed a family support organiser (FSO) service and refined outcome measures for evaluation. We tested the effects of the intervention in a randomised controlled trial. Methods: From 1 March 1999 to 1 April 2001 all first-in-a-lifetime strokes (n = 513) were identified and 340 (96%) of eligible strokes randomised to receive FSO or usual care. Patients and their carers were followed up at 3 months and 1 year post-stroke. Outcomes included satisfaction (main outcome) with hospital staff and outpatient services, use of social services, reintegration to normal living (RNLI) and feelings about life after the stroke. Results: The mean number of contacts with the FSO was 15 (SD = 9.8) per patient. More intervention than control patients received some social services and had increased patient and carer satisfaction in most aspects, particularly with information about recovery and feeling that someone had listened. There was little evidence at 3 or 12 months of differences in RNLI. Conclusions: A meta-analysis of trials in this area is now needed along with further trials of interventions in subgroups of the stroke population to fully identify any benefits of the FSO role.