In 84 patients (85 elbows), a standard goniometer was used to measure elbow flexion and extension with the forearm in 2 positions: full supination (humeral condyles parallel to floor) and neutral forearm rotation (ulnohumeral joint in relative internal rotation). All measurements were made by the same surgeon with a standardized technique. Mean absolute difference in elbow extension measurements was 8°. There was no difference in flexion between measurement positions. Mean difference between neutral and supinated measurements was 6° in extension. There was no difference in flexion for patients with contractures between 0° and 29°. In patients with flexion contractures of more than 30°, mean difference between neutral and supinated measurements was 12° in extension; there was no difference in flexion. Elbow goniometer extension measurements taken in neutral humeral rotation underestimated the degree of elbow hyperextension and elbow flexion contracture. We recommend taking elbow flexion arc measurements in the true plane of motion, with the humerus externally rotated by fully supinating the forearm, such that the distal humeral condyles are parallel to the floor.