Objectives: There is inconclusive evidence of the effectiveness of stroke rehabilitation by a community stroke team. The aim was to evaluate a specialist multiprofessional team in a community setting. Design: Randomized controlled trial. Setting: Community. Participants: Stroke patients and their informal carers who were referred to receive rehabilitation from a community stroke team. Outcome measures: Barthel Index, Extended Activities of Daily Living Scale (EADL), General Health Questionnaire (GHQ-12) by patient and carer, Carer Strain Index (CSI), Euroquol, knowledge of stroke and satisfaction with services six months after recruitment. Results: There were no significant differences between patients who received rehabilitation from community stroke team ( n = 189) and those who received routine care ( n = 232) in their independence in activities of daily living, mood, quality of life or knowledge of stroke. The patients in the community stroke team group were significantly more satisfied with the emotional support they had received ( p < 0.01). There were no significant differences between the groups in satisfaction with practical help or overall satisfaction. Carers of patients in the community stroke team were under significantly less strain than carers in the routine care group ( p < 0.04). Carers of patients in the community stroke team group were significantly more satisfied with their knowledge of stroke ( p < 0.01) and were more satisfied overall ( p < 0.01). Conclusions: The patients treated by the community stroke team were more satisfied with the emotional support they received and had equivalent outcomes in terms of independence in activities of daily living and mood. Their carers were under less strain and were more satisfied with their knowledge of stroke recovery, the emotional support they received and overall satisfaction with services. The results support the provision of rehabilitation by a community-based specialist multiprofessional team.