The aim of this study was to investigate the kinematics of young adults during descent ramp climbing at different inclinations. Twenty-three subjects descended four steps at four different inclinations (level, $-8^{circ}$, $-16^{circ}$, $-24^{circ}$). The 3-D kinematics were measured by a camera-based Falcon System. The data were analyzed using one-way ANOVA and the Student-Newman-Keuls test. The kinematics of descent ramp walking could be clearly distinguished from the kinematics of level walking. On a sagittal plane, the ankle joint was more plantar flexed at initial contact with $-16^{circ}/-24^{circ}$ inclination, was decreased in the toe off position with all inclinations (p<.001),and was decreased at maximum plantar flexion during the swing phase (p<.001). The knee joint was more flexed at initial contact with the $-24^{circ}$ inclination (p<.001), was more flexed in the toe off position with all inclinations (p<.001), and was more flexed at minimum flexion during stance phase and at maximum flexion during swing phase with $-16^{circ}$, $-24^{circ}$ inclination (p<.001). The hip joint was more flexed in the toe off position with $-16^{circ}$, $-24^{circ}$ inclination and was deceased at maximum extension during stance phase with $-16^{circ}$, $-24^{circ}$ inclination (p<.05). In the frontal plane, the ankle joint was more everted at maximum eversion during stance phase with $-16^{circ}/-24^{circ}$ inclination (p<.01) and was decreased at maximum inversion during swing phase with $-16^{circ}$, $-24^{circ}$ inclination (p<.01). The knee joint was more increased at maximum varus during stance phase with $-16^{circ}/-24^{circ}$ inclination (p<.001). The hip joint was deceased at maximum adduction during stance phase with $-24^{circ}$ inclination (p<.05). In a horizontal plane, only the knee joint was increased at maximum internal rotation during stance phase with $-24^{circ}$ inclination (p<.05). In descent ramp walking, the different gait patterns occurred at an inclination of over $16^{circ}$ on the descending ramp in the sagittal and frontal planes. These results suggest that there is a certain inclination angle or angular range where subjects do switch between level walking and descent ramp walking gait patterns.