A typical tennis match can take 1 to 5 hours. Tennis serve can be divided into eight stages and three phases. Injuries usually happen in the cocking stage and acceleration phase of the serve, where it requires large shoulder range of motion and proper scapular motion, including sufficient scapular upper rotation, external rotation, and posterior tilt to produce a powerful serve. Altered shoulder kinematics are associated with shoulder injuries in tennis players, including delayed shoulder horizontal adduction, and early external rotation. Due to the long duration and repetition of activity in a match, fatigue may happen and result in decreases in sensory input, passive range of motion, ball speed, and muscle strength. These changes may further lead to altered glenohumeral and scapular kinematics. However, previous studies mainly focused on the effects of fatigue on scapular kinematics in constrain movements and applied different fatigue protocols, which lead to inconsistent results. Tennis serve in cocking stage and the acceleration phase require high activation of shoulder external rotators and internal rotators, including infraspinatus, pectoralis major, subscapularis, latissimus dorsi and serratus anterior. However, to our knowledge, no study has investigated how fatigue of shoulder rotators influences shoulder kinematic as well as scapular kinematics during the late cocking stage and acceleration phase of tennis serve. Therefore, the investigators aim to investigate whether fatigue of shoulder rotator affects shoulder kinematics in healthy tennis players during the late cocking stage and acceleration phase of tennis serve.This is a single group, pretest-posttest measurement study. In a fatigue protocol, investigators use an isokinetic dynamometer to induce fatigue of shoulder rotators. Outcome measures will be tested before and after the fatigue protocol, including peak torque of shoulder rotators, humerothoracic kinematics, scapulothoracic kinematics, and median power frequency recorded by a surface electromyography. Peak torque of shoulder rotators will be measured with an isokinetic dynamometer. Surface electromyography will be used to measure peripheral muscle fatigue by maximum voluntary isometric contraction. Humerothoracic kinematics and scapulothoracic kinematics during a functional tennis serve and scaption will be collected with a motion capture system.

Inclusion Criteria:

Healthy tennis players

1. Age ranges from 20-45 years old
2. Engaging in tennis for at least 1 year
3. Routine training 3 times a week
4. No history of cervical and/or shoulder injury in the past one year, or injury with VAS (visual analog pain scale) lower than 3 and without cervical-shoulder pain 7 days before testing
5. No operation on the shoulder, no shoulder subluxation, labral tear, and rotator cuff tear.
6. Neer's test, empty can test, and Hawkins Kennedy test are negative.

Exclusion Criteria:

1. History of cervical and/or shoulder injury in the past one year, or injury with VAS (visual analog pain scale) bigger than 3 and with cervical-shoulder pain 7 days before testing
2. Operation on the shoulder, history of shoulder subluxation, labral tear, and rotator cuff tear.
3. One of the three test results is positive: Neer's test, empty can test, and Hawkins Kennedy test.
4. Perform upper extremity resist training and tennis training 24 hours before testing.

All of the subjects are healthy during the test to avoid the risk of injury.