Recent reports link GLP-1RA treatment with a higher risk of non-arteritic anterior ischemic optic neuropathy, especially in the first year, but the underlying mechanisms are unclear. This translational study investigated GLP-1RA’s impact on retinal perfusion and the mediating role of intracranial pressure (ICP) using clinical observations and a mouse model. In a small retrospective cohort, fluorescein angiography revealed a trend toward prolonged arteriovenous transit time (AVTT) during active GLP-1RA treatment, which normalized post-treatment. Consistently, GLP-1RA-treated mice exhibited prolonged arterial appearance and AVTT, accompanied by diminished scotopic b-wave responses and upregulated retinal hypoxia markers. Mechanistically, these mice showed significantly reduced ICP and retinal venous diameter on day 7, corresponding with the timing of retinal perfusion delays. Significant correlations were observed between day 7 ICP and AVTT changes ( R² = 0.7605, p  = 0.0010), and between day 7 AVTT and venous diameter changes ( R² = 0.4247, p  = 0.0154). Based on these findings, we hypothesize that GLP-1RA-induced ICP reduction may lead to mechanical alterations in the subarachnoid space—such as narrowing or collapse of the optic nerve sheath—which subsequently compress the central retinal vessels and compromise arterial inflow. These hemodynamic alterations in mice were independent of systemic cardiovascular changes and were fully reversed upon treatment cessation, mirroring the transient trends observed in humans. Collectively, our results suggest that the ocular circulation may be sensitive to GLP-1RA-associated hemodynamic changes, potentially mediated by ICP reduction. Further prospective clinical investigation into the temporal relationship between ICP and retinal hemodynamics in GLP-1RA users is warranted.