Purpose The aim of this study was to explore the cerebral distribution of the tau-specific PET tracer [F-18]THK5317 (also known as (S)-[F-18]THK5117) retention in different stages of Alzheimer's disease; and study any associations with markers of hypometabolism and amyloid-beta deposition. Methods Thirty-three individuals were enrolled, including nine patients with Alzheimer's disease dementia, thirteen with mild cognitive impairment (MCI), two with non-Alzheimer's disease dementia, and nine healthy controls (five young and four elderly). In a multi-tracer PET design [F-18]THK5317, [C-11] Pittsburgh compound B ([C-11]PIB), and [F-18]FDG were used to assess tau pathology, amyloid-beta deposition and cerebral glucose metabolism, respectively. The MCI patients were further divided into MCI [C-11]PIB-positive (n=11) and MCI [C-11]PIB-negative (n=2) groups. Results Test-retest variability for [F-18]THK5317-PET was very low (1.17-3.81 %), as shown by retesting five patients. The patients with prodromal (MCI [C-11]PIB-positive) and dementia-stage Alzheimer's disease had significantly higher [F-18]THK5317 retention than healthy controls (p=0.002 and p=0.001, respectively) in areas exceeding limbic regions, and their discrimination from this control group (using the area under the curve) was >98 %. Focal negative correlations between [F-18]THK5317 retention and [F-18]FDG uptake were observed mainly in the frontal cortex, and focal positive correlations were found between [F-18]THK5317 and [C-11] PIB retentions isocortically. One patient with corticobasal degeneration syndrome and one with progressive supranuclear palsy showed no [C-11]PIB but high [F-18]THK5317 retentions with a different regional distribution from that in Alzheimer's disease patients. Conclusions The tau-specific PET tracer [F-18]THK5317 images in vivo the expected regional distribution of tau pathology. This distribution contrasts with the different patterns of hypometabolism and amyloid-beta deposition.