Rater reliability characteristics of four positioning techniques used to measure hip extension (prone hip extension, Thomas, Mundale, and pelvifemoral angle) were compared. Two raters examined 45 children (90 hips), including patients with spastic diplegia and with meningomyelocele, who are prone to developing hip flexion contractures, and healthy subjects. Retests were performed on 21 children. The least reliable readings in the cerebral palsy group were found from tests using the Thomas technique, and in the meningomyelocele group using the Mundale technique. The pelvifemoral angle is more time consuming, and no more accurate than other methods. Considering the ease of measurement, reliability and reproducibility, we recommend use of the prone hip extension test for patients with cerebral palsy and for those with meningomyelocele, and recommend the Thomas test as an alternative for nonspastic patients.