© 2018 Japan Geriatrics Society Aim: Rates of disease progression differ among patients with Alzheimer's disease (AD), but prognostic predictions remain a challenge. We carried out a clinic-based retrospective study to investigate the clinical factors for AD progression. Methods: The 748 AD patients, who attended our hospital for >1 year and were given the Mini-Mental State Examination (MMSE) at least three times, were divided into three groups according to the annual change rate of MMSE score (G): Aggravater group (G < −2), Stabler group (−2 ≤ G ≤ 2) and Improver group (2 < G). We compared the three groups on cognitive, affective and activities of daily living functions, response to medication, clinical fluctuations, serum levels of metabolic factors, and neuroimaging data. Results: We found no significant differences in age, sex, educational attainment or body mass index across the groups. The Aggravater group showed better baseline MMSE (P < 0.01) and Abe's behavioral and psychological symptoms of dementia (P < 0.01) scores than the Improver group, but its MMSE improvement after drug treatment was the worst among the three groups (P < 0.01 vs Stabler/Improver). Fluctuations in MMSE (P < 0.01), apathy scale (P < 0.05) and activities of daily living (P < 0.01) scores were smaller in the Improver group than in the Aggravater or Stabler groups. Serum docosahexaenoic acid levels tended to be lower (trend P < 0.05) and voxel-based specific regional analysis system for Alzheimer's disease Z-scores tended to be higher (trend P < 0.05) in the Improver group than in the Stabler or Aggravater groups. Conclusions: Initial responses to medication, fluctuations in cognitive, affective and activities of daily living functions, serum docosahexaenoic acid levels, and medial temporal atrophy are clinical factors related to AD prognosis. Geriatr Gerontol Int 2018; 18: 929–936.