Two series of healthy men were measured for range of passive motion of hip flexion, hip extension, hip abduction, knee flexion, and ankle dorsiflexion with the knee extended and flexed. Hip abduction was measured with a specially constructed double protractor goniometer, and the other movements were gauged with a flexometer. In series A measurements were based upon a commonly used clinical method. The intratester interassay coefficient of variation in series A was 7.5 +/- 2.9%, which corresponded well with other reports. In series B posture and measuring procedures were rigidly standardized by better fixation and by identification and marking of anatomical landmarks. The interassay coefficient of variation in series B was reduced to 1.9 +/- 0.7%. Range of motion measurement was repeated accurately by the same tester with methods requiring careful measurement technique but no elaborate equipment.