Primary open-angle glaucoma (POAG) is now seen as a progressively worsening neurodegenerative disorder. Previvors neuroimaging studies neglect dynamic brain activity, assuming static patterns in resting-state functional MRI (rs-fMRI), while intrinsic brain activity is dynamic and fluctuates over time. To test the time-varying brain activity in POAG patients, a temporal dynamic analysis of rs-fMRI data was conducted on 70 POAG patients and 45 healthy controls (HCs). The sliding-window method calculated dynamic amplitude of low-frequency fluctuations (dALFF) across the brain, and group differences and correlations with ophthalmological and neuropsychological measures were analyzed. Multivariate pattern analysis evaluated dALFF and static-ALFF's ability to differentiate POAG patients from HCs. POAG patients exhibited lower dALFF in visual network areas and higher dALFF in the default mode, ventral attention, and frontoparietal networks compared to HCs. Decreased dALFF in the cuneus correlated with visual field deficits. Visual network dALFF distinguished POAG patients from HCs with 79.13 % accuracy, 66.67 % sensitivity, and 87.14 % specificity (p = 0.001). The dALFF's area under the curve was significantly greater than that of static-ALFF (p = 0.026). POAG patients show irregular time-varying brain activity, especially in areas linked to vision, movement, and emotional/cognitive functions. This study suggests that dALFF offers a novel approach to capture the aberrant time-varying brain activity and its clinical relation, holding potential for enhancing the understanding of POAG mechanisms and improving its diagnosis.