Background: Type 2 diabetes mellitus (T2DM) remains one of the greatest threats to vision globally. HbA1c is the gold standard biomarker of blood glucose control, but recent studies have begun to explore the role of Glycemic variability (GV) as a complementary metric capable of stratifying patients’ risk of complications independent of HbA1c. Studies have linked greater GV to reduced retinal microvasculature density in type 1 diabetes mellitus (T1DM), potentially a pathophysiological result of increased GV leading to blood vessel damage. However, no studies have examined the effects of GV on the retinal microvasculature in T2DM. Optical coherence tomography angiography (OCTA) has gained acceptance in Ophthalmology to reproducibly quantify the retinal microvasculature. Here, we sought to correlate quantitative retinal OCTA metrics to GV. Methods: Cohort study of adult patients with T2DM. We used the publicly available AI READI dataset, which as of July 2024 includes 204 T2DM patients across three data collection sites and contains clinical characteristics, retinal OCTA imaging (Maestro2, Topcon, Tokyo, Japan), and continuous glucose monitoring (CGM; Dexcom G6, Dexcom, San Diego, USA), among other data. OCTA metrics (vessel area density (VAD) and vessel length density (VLD)) were calculated for the superficial (SCP) and deep (DCP) capillary plexuses of each eye of each patient, and patients with VAD or VLD less than 10% or 10mm^-1, respectively, were excluded for low image quality. The coefficient of variation (CoV) of CGM-measured plasma blood glucose recorded every 5 minutes over 10 days was calculated. Patients with a CoV greater than 0.15 for a particular metric were classified as high variation for that metric, and the remainder as low variation. Difference in mean values between high variation and low variation groups were examined for each OCTA metric (unpaired t-tests). Pearson correlation coefficients were calculated for each calculated OCTA metric and the CoV of CGM-measured blood glucose. Results: 198 patients were identified with available imaging. Of these 24 were excluded for poor image quality. A statistically significant difference in VAD between high and low-variation patients was identified, with those with higher variability having lower VAD (p<0.05). Conclusion: This study extends prior results finding a correlation between higher blood glucose variability and lower retinal microvascular density from T1DM to T2DM. Blood glucose variability may play a significant role in the pathophysiology of microvascular disease in diabetes.