Background The tennis serve is the most common stroke in tennis and is the single stroke most associated with musculoskeletal injury in tennis. Previous investigations have utilised marker-based systems which are prone to error and also have not evaluated all three types of serves.

Objective Utilise a novel markerless video system to determine whether the kick and slice serves produced greater forces/torques about the torso and upper extremity versus the flat serve.

Design Single recording session.

Setting Outdoor regulation tennis court.

Participants Seven NCAA Division I male tennis players.

Interventions Three successful flat, kick, and slice serves.

Main outcome measurements Back extension angle; total forces at back, shoulder, elbow, and wrist.

Results Maximum back extension angle was significantly higher for the kick serve versus the slice (40.5 vs 37.3°; p=0.01) while there was a trend towards significance in the kick compared to the flat serve (40.5 vs 38.6°; p=0.096). Maximum external rotation of the shoulder was not significantly different between flat, kick, and slice serves (89.8 vs 90.1 vs 91.5°; p>0.67). The kick serve showed greater total back force (2974 vs 2138 vs 2568 N; p<0.05) and posterior shoulder force (1370 vs 809 vs 741 N, p<0.05) versus the flat and slice. The slice serve demonstrated lower torques at the elbow (71.1 vs 87.0 vs 89.3 N-m; p<0.05) and wrist (20.9 vs 33.5 vs 27.9 N-m; p<0.05) compared to the flat and kick serves.

Conclusion The kick serve places higher stresses on the back and shoulder while the slice is the least injurious to the elbow and wrist. This data may help provide guidelines for injury prevention, rehabilitation, and possibly for return to play after injury or surgery.