Abstract Disclosure: S. Syed: None. N. Binkley: None. Introduction: Prevalent vertebral fractures convey increased fracture risk but are often clinically silent; thus, imaging is required for their detection. Vertebral fracture assessment (VFA) can be performed at the time of DXA. We hypothesized that VFA automated morphometry does not accurately identify the number of vertebrae visualized nor the number of fractures present and sought to begin exploring if morphometry impacted the formal interpretation. Methods: Patients seen by an Orthopedic bone health provider were studied. Their VFA images were independently assessed by two physicians, after which a consensus “gold standard” (GS) reading was reached. The number of vertebral bodies able to be assessed and the number of fractures identified by morphometry (M), the formal interpretation (I) and the GS result were compared by one way ANOVA. Results: The study cohort included 100 patients with mean (SD) age 77.6 (8.6) years, 81% female with BMI 26.7 (5.0) kg/m2, 91 of whom had prior clinical fracture. M reported visualizing more vertebral bodies than I and GS; 1243, 1168 and 1063 respectively. I identified numerically more (13) fractures than did M (150 vs. 137) but was lower than GS (167). The automated and formal reports were identical in 52 patients; in this subgroup, 8 patients (15%) had vertebral fractures that were not identified on the formal interpretation. In this same group of 52 patients, an additional 7 (13%) with vertebral fractures had 9 vertebral fractures missed. Conclusion: In this pilot study assessing VFA, automated vertebral morphometry reports identifying more vertebrae than are actually able to be evaluated while missing almost 18% of vertebral fractures that were actually present. That the formal interpretation is identical to that of morphometry in over 50% of patients, while missing vertebral fractures over 12% of the time, implies that morphometry is adversely impacting the formal interpretation and thereby potentially patient care. Further study with larger datasets is indicated. Presentation: 6/3/2024