PURPOSE: The measurement of ocular torsion remains complicated and delicate when using subjective methods of measurement. Objective measures are now accessible in daily practice. This study determined the standards for ocular torsion in healthy patients with fundus photographs. MATERIALS AND METHODS: Prospective study on 150 patients divided into three different age groups: children (under 15 years); young adults (15-50 years); and adults (over 50 years). All patients included had a normal oculomotor examination and were close to emmetropia. The fundus photographs were taken using a standardized system. The results were analyzed with a graphic computer program and took into consideration reproducible anatomical models; a statistical study determined the thresholds of statistical significance. RESULTS: The optic nerve head-fovea angle followed a Gaussian distribution (mean, 6.3; standard deviation, 3.4). The measure was reproducible in a single patient through different successive examinations. There was no variation of this angle between the patients of different ages. Furthermore, in 87% of the patients, the fovea had a projection between the center and the lower edge of the optic nerve head, indicating that these anatomical models are reliable in a clinical evaluation of torsion. DISCUSSION: This study confirms the accuracy of the objective measurements of ocular torsion and validates the anatomical models proposed by the rare studies already published. CONCLUSIONS: The era of systematic ophthalmic photography is opening new perspectives in diagnosis now that the range in healthy patients has been determined, necessary before the disease could be analyzed. This validation of reliable and simple anatomical models should make this method easy to use in clinical and daily practice.