Objectives: To compare the 4-year survival, institutionalization, cognitive and functional decline of Alzheimer's patients with specific follow-up in memory centers versus usual care. Design: Four year longitudinal follow-up. Settings: The French Network of memory centers in Alzheimer's disease (REAL-FR study) and The French population-based study (3C study). Participants: 728 patients aged ≥ 65, living at home, meeting criteria for probable Alzheimer's disease and having Mini Mental State Examination (MMSE) scores between 10 and 26 at baseline were included. Measurements: Cox proportional hazards models were performed to test the effectiveness of a specific follow-up in memory centers (REAL-FR study) versus usual care (3C study) on the 4-year survival and institutionalization. Linear mixed models were used to assess cognitive and functional decline in both groups. Results: After adjustment for confounding factors, the 4-year survival did not differ significantly between patients followed-up in memory centers and those who had recourse to usual care (usual care: Hazard Ratio adjusted (HRa) = 0.87, 95% confidence interval (CI) 0.53-1.43, p=0.59). Patients with a specific follow-up in memory centers had a higher risk of being institutionalized (usual care: HRa = 0.24, 95% CI 0.12-0.48, p<0.001). They also exhibited a significant greater cognitive and functional decline over time. Conclusion: Our findings failed to demonstrate any potential benefits of a specific follow-up in memory centers on clinically meaningful outcomes in the natural history of Alzheimer's disease. Recourse to care in memory centers may have been the consequence of a faster dementia progression and a greater burden of Alzheimer's disease, all leading to detrimental consequences on various prognostic outcomes.