The purpose of this study was to investigate the relationship between the maximum angular velocities at the shoulder and elbow joints and the consistency of the cycle-movement pattern in wheelchair racing. Eleven elite international athletes competing in the 1999 Roosevelt International Cup in Warm Springs, Georgia, participated in this study. Their mean age was 30 years old, (standard deviation 7.8). Their wheelchair racing experiences ranged from 2 to15 years. A ProReflex Imaging System with two cameras was employed to film the wheelchair propulsion cycles on a roller with athletes' own chairs. Participants performed all-out speeds in wheelchair propulsion. Five reflective markers placed on the shoulder, elbow, and wrist joints, on the metacarpal and wheel axis were automatically digitized. Three trials (7 seconds per trial) were recorded at 100 Hz. Ten continuous cycles in a trial were selected for analysis. The cycle time, the angular ranges of motion (AROM) and angular velocities of the shoulder and elbow joints were examined. Then, the deviations of cycle time, joint angle positioning, AROM and angular velocity among the cycles were calculated. Pearson-Product correlation analysis (p<0.05) was used to determine the relationship between the cycle-movement deviations and the angular velocities. The results of the significant correlation coefficients were that at the shoulder joint: the deviation of joint positioning to the deviation of angular velocity (r=0.881), to the AROM (r=0.724), and to the maximum velocity (r=0.808), and the AROM to the maximum velocity (r=0.935); at the elbow joint: the deviation of joint angle positioning to the deviation of angular velocity (r=0.973), to the deviation of cycle time (r=0.757), the deviation of angular velocity to the deviation of cycle time (r=0.803), and the AROM to the maximum angular velocity (r=0.845). The results demonstrated that the more deviations in cycle time and joint angle positioning were associated with the greater angular velocities at the shoulder and elbow joints for the elite wheelchair racers. The interpretations may be that the less consistency of cycle-movement pattern in wheelchair racing might be associated with alternately firing different motor units of the muscles which would facilitate the generation of maximum angular velocity. Furthermore, the consistency of the cycle-movement pattern and the efficiency are not the concern in sprint wheelchair racing. Future study is needed to verify the muscles' motor units firing patterns at the shoulder and elbow joints during wheelchair racing. Keyword(s): athletics/sports, physical activity, research