Objectives Evaluation of both the diagnostic accuracy of ultrasound, and of Emergency Medicine (EM) physicians, for pediatric ovarian torsion (OT). Methods This study consists of a retrospective review of 62 female pediatric patients presenting to a tertiary care setting, diagnosed with OT, over a 10-year period. Reviewers extracted radiographic data from the patients identified, including ultrasonographic data supporting OT and comparing diagnosis based on ultrasound versus definitive laparoscopy. Statistical analysis included calculations of sensitivity/specificity, positive predictive value and t tests. Results In patients identified as being at risk for OT, sensitivity and specificity were respectively 85.7% and 62.9% for ultrasound, and 80.0% and 57.7% for Doppler flow. In a comparison of patients affected to unaffected ovary, we found a significant size difference across all three dimensions measured: 5.54cm to 2.57cm, 5.1cm to 2.31cm and 4.11cm to 2.38cm. 51% of EM referrals for query OT were truly torsion, while 81% of patients taken to the operating room by the Obstetrical/Gynecology consultants had a true torsion. Notably, seven patients were misdiagnosed in the emergency department, and found to have a final diagnosis of OT. This supports our hypothesis that OT is necessarily a largely clinical diagnosis. Conclusions We revealed that within our center, pediatric EM PPV for OT was 0.51, which is increased to 0.81 when gynecology performs a laproscopy, highlighting the diagnostic dilemma for clinicians in capturing this diagnosis. Ultrasound alone produced moderate sensitivity, making it insufficient to rule out OT, supporting the SOGC's guideline that this is a clinical diagnosis.