Cyclooxygenase (COX), a prostanoid-synthesizing enzyme, is considered to be involved in the neuroinflammatory process of neurodegenerative diseases. However, the role of COX in the progression of neurodegeneration is not well understood. We hypothesized that in vivo imaging of COX by PET will contribute to elucidation of the function of COX during the neurodegenerative process in Alzheimer's disease (AD). 11 C-labeled ketoprofen methyl ester (racemic (RS)-11 C-KTP-Me) developed recently by our group is a useful PET probe for in vivo imaging of COX-1 during neuroinflammation. The (S)-enantiomer of ketoprofen is known to be pharmacologically more active than the (R)-enantiomer. We thus synthesized 11 C-labeled (S)-ketoprofen methyl ester ((S)-11 C-KTP-Me) as an improved PET probe specific for COX-1 and applied it for investigation of the changes in COX-1 during the progression of AD in a mouse model. Methods: The specificity of (S)-11 C-KTP-Me for COXs was examined in PET studies with rats that had intrastriatal injection of lipopolysaccharide. To determine the details of changes in COX-1 during progression of amyloid-β (Aβ) plaque formation in amyloid precursor protein transgenic (APP-Tg) mice, we performed immunohistochemical studies and ex vivo autoradiography with (S)-11 C-KTP-Me. Results: PET studies using hemispheric lipopolysaccharide injection into rats revealed that the sensitivity of (S)-11 C-KTP-Me in neuroinflammation was much higher than that of (RS)-11 C-KTP-Me and (R)-11 C-KTP-Me; these results closely corresponded to the inhibitory activities of each enantiomer against COX-1 estimated by an in vitro assay. In APP-Tg mice, (S)-11 C-KTP-Me administration resulted in progressive and significant increases in accumulation of radioactivity in the brain from 16 to 24 mo old in accordance with the histopathologic appearance of abundant Aβ plaques and activated microglia, whereas few changes in radioactivity accumulation and few Aβ plaques were seen in age-matched wild-type control mice. High-radioactivity accumulation by (S)-11 C-KTP-Me was markedly observed in the frontal cortex and hippocampus in which COX-1-expressing activated microglia tightly surrounded and enclosed large and more intensely stained Aβ plaques, indicating neuroinflammation that originated with Aβ. Conclusion: (S)-11 C-KTP-Me is a potent PET probe that is highly selective for COX-1. Studies using APP-Tg mice demonstrated that (S)-11 C-KTP-Me could detect activated microglia that are associated with amyloid plaque progression, suggesting the involvement of COX-1 in the neuroinflammatory process in AD.