Objective: To assess whether shared care for stroke patients results in better patient outcome, higher patient satisfaction and different use of healthcare services. Design: Prospective, comparative cohort study. Setting: Two regions in the Netherlands with different healthcare models for stroke patients: a shared care model (stroke service) and a usual care setting. Patients: Stroke patients with a survival rate of more than six months, who initially were admitted to the Stroke Service of the University Hospital Maastricht (experimental group) in the second half of 1997 and to a middle sized hospital in the western part of the Netherlands between March 1997 and March 1999 (control group). Main outcome measures: Functional health status according to the SIP-68, EuroQol, Barthel Index and Rankin Scale, patient satisfaction and use of healthcare services. Results: In total 103 patients were included in this study: 58 in the experimental group and 45 in the control group. Six months after stroke, 64% of the surviving patients in the experimental group had returned home, compared to 42% in the control group (p<0.05). This difference could not be explained by differences in health status, which was comparable at that time. Patients in the shared care model scored higher on patient satisfaction, whereas patients in the usual care group received a higher volume of home care. Conclusions: The Stroke Service Maastricht resulted in a higher number of patients who returned home after stroke, but not in a better health status. Since patients in the usual care group received a higher volume of healthcare in the period of rehabilitation, the Stroke Service Maastricht might be more efficient.