Objective: To evaluate the effect of an extended stroke unit service (extended service), with early supported discharge and co-ordination of further rehabilitation in co-operation with the primary health care system in three rural municipalities. Design: A randomized controlled trial comparing extended service with ordinary stroke unit service (ordinary service). Subjects: Sixty-two eligible patients with acute stroke living in the rural municipalities of Malvik, Melhus and Klñ bu. Main measures: The primary outcome was the proportion of patients who were independent according to Modified Rankin Scale (mRS) (independence = mRS < 2) 52 weeks after onset of stroke. Secondary outcomes were mRS at 6 and 26 weeks and Barthel Index (BI), Nottingham Health Profile (NHP) and Caregiver Strain Index (CSI) at 6, 26 and 52 weeks. Mortality and length of stay were registered during the 52 weeks. Results: Twelve patients (39%) in the extended service group versus 16 patients (52%) in the ordinary service group were independent according to mRS at 52 weeks (p= 0.444). The odds ratio for independence (extended service versus ordinary service) was 0.33 (95% confidence interval (CI) 0.088 –1.234). According to outcome by secondary measures there were no significant differences except less social isolation on NHP in the extended service group at 26 weeks (p= 0.046). There were no significant differences in length of stay. Conclusion: An extended stroke unit service with early supported discharge seems to have no positive effect on functional outcome for patients living in rural communities, but might give a trend toward better quality of life. There were no significant differences in length of stay.