PURPOSE: To evaluate the reliability and clinical usefulness of ultrasonography in the temporomandibular joint (TMJ). MATERIALS AND METHODS: Parasagittal and paracoronal 1.5 T MR images and 7.5 MHz ultrasonographs of 40 TMJs in 20 asymptomatic volunteers were obtained. Disc position using MR imaging was evaluated and the distance between the lateral surface of mandibular condyle and the articular capsule using MR image and ultrasonograph of 27 TMJs with normal disc position was measured and compared. Intraobserver and interobserver measurements reliability was evaluated by using interclass correlation coefficients (ICC) and measurement error. Also, the distance measured on ultrasonographs was compared, according to mouth position and disc postion. RESULTS: The normal disc position was found in 27 of 40 asymptomatic joints. At the intraobserver reliability of measurement, ICC at the closed and open mouth position were 0.89 and 0.91. The measurement error was 0.4% and 0.5%. At the interobserver reliability, ICC at the closed and open mouth position were 0.92 and 0.81. The measurement error was 0.4% and 0.7%. At the TMJ with normal disc position, the distances between the lateral surface of mandibular condyle and the articular capsule measured on MR images and ultrasonographs were 2.0+/- 0.7 mm, 1.8+/-0.5 mm, respectively (p<0.05). On the ultrasonographs, the distances at open mouth position were 1.2+/-0.5 mm (p<0.05). At the TMJ with medially displaced disc, the distances at the closed and open mouth position were 1.3+/-0.3 mm and 0.9+/-0.2mm(p<0.05). CONCLUSION: The results suggest ultrasonography of TMJ is a reliable imaging technique for assessment of normal disc position.