Abstract Purpose We assess the clinical accuracy of direct-to-patient real-time outpatient video visit encounters at our eye center. Design This was a retrospective longitudinal study. Subjects and Methods Patients who completed a video visit over a 3-week period between March and April 2020 were included. Accuracy assessment was determined by comparing diagnosis and management from the video visit with subsequent in-person follow-up over the next year. Results A total of 210 patients (mean age 55±18 years) were included, of whom 172 (82%) were recommended a scheduled in-person follow-up encounter after their video visit. Among the 141 total patients who completed in-person follow-up, 137 (97%) had a diagnostic agreement between telemedicine and in-person evaluation. Management plan agreed for 116 (82%), with the remainder of visits either escalating or deescalating treatment upon in-person follow-up with little substantive change. Compared with established patients, new patients had higher diagnostic disagreement following video visits (12 vs. 1%, p=0.014). Acute visits trended toward more diagnostic disagreement compared with routine visits (6 vs. 1%, p=0.28) but had a similar rate of management change on follow-up (21 vs. 16%, p=0.48). New patients were more likely to have early unplanned follow-up than established patients (17 vs. 5%, p=0.029), and acute video visits were associated with unplanned early in-person assessments compared with routine video visits (13 vs. 3%, p=0.027). There were no serious adverse events associated with the use of our telemedicine program in the outpatient setting. Conclusions Video visits had high diagnostic and management agreement with subsequent in-person follow-up encounters.