Proprioception in the neck was investigated in normal human subjects. Three experiments studied rotation of the head about a vertical axis on the body. Accuracy of pointing, thresholds for detection of passive movement, and control of fine movement were tested. Comparison of the accuracy of pointing at the big toe with the nose and with the arm, showed a smaller scatter of angular misalignments when pointing with the arm. However, the arm pointed systematically off target. Pointing at the target toe by turning the head was not significantly more accurate than aligning the nose and toe by turning the chair and body with the head fixed. The highest threshold found for the detection of the direction of passive movement of the head relative to the body was 1.4 degrees angular displacement. Thresholds were highest at the slowest angular velocity and dropped as angular velocity increased. When the head was turned on the body thresholds were lower than when the body was turned and the head held still. Control of fine angular movements of the head and of the distal phalanx of the right thumb were compared by measuring subjects' accuracy in guiding a cursor through a path on a computer screen by turning the head or moving the thumb. The thumb was found to be better controlled than the head.