Background: Tibial torsion is a twist in the tibia measured as an angle between a proximal axis line and a distal axis line. Abnormal torsion has been associated with a variety of painful clinical syndromes of the lower limb. Measurements of normal tibial torsion reported by different authors vary by 100% (ranging from 20° to 42°), making it impossible to determine normal and pathological levels. Purpose: To address the problem of unreliable measurements, this study was conducted to define an automated, validated computer method to calculate tibial torsion. Reliability was compared with current clinical methods. The difference between measurements of torsion generated from computed tomography (CT) and magnetic resonance imaging (MRI) scans of the same bone, and between males and females, was assessed. Study Design: Controlled laboratory study. Methods: Previous methods of analyzing tibial torsion were reviewed, and limitations were identified. An automated measurement method to address these limitations was defined. A total of 56 cadaveric and patient tibiae (mean ± SD age, 37 ± 15 years; range, 17-71 years; 28 female) underwent CT scanning, and 3 blinded assessors made torsion measurements by applying 2 current clinical methods and the automated method defined in the present article. Intraclass correlation coefficient (ICC) values were calculated. Further, 12 cadaveric tibiae were scanned by MRI, stripped of tissue, and measured using a structured light (SL) scanner. Differences between torsion values obtained from CT, SL, and MRI scans, and between males and females, were compared using t tests. SPSS was used for all statistical analysis. Results: When the automated method was used, the tibiae had a mean external torsion of 29°± 11° (range, 9°-65). Automated torsion assessment had excellent reliability (ICC, 1), whereas current methods had good reliability (ICC, 0.78-0.81). No significant difference was found between the torsion values calculated from SL and CT (P = .802), SL and MRI (P = .708), or MRI and CT scans (P = .826). Conclusion: The use of software to automatically perform measurements ensures consistency, time efficiency, validity, and accuracy not possible with manual measurements, which are dependent on assessor experience. Clinical Relevance: We recommend that this method be adopted in clinical practice to establish databases of normal and pathological tibial torsion reference values and ultimately guide management of related conditions.