BACKGROUND: Low back pain (LBP) is pervasive among elite junior tennis players. Previous research has explored the relationship between serving mechanics and LBP, though the participants in these studies had already experienced LBP. Therefore, it is unclear whether their serving mechanics caused the LBP or are a result of having LBP. Thus, the purpose of this study was to compare the flat and kick serve kinematics of asymptomatic elite adolescent male and female tennis players with and without lumbar spine abnormalities. Twenty-four players (nine of which had confirmed lumbar spine abnormalities) carried out a series of flat and kick serves, while marker trajectories were recorded by a 3D motion capture system. Pelvis and lumbar spine kinematics (anterior/posterior tilt, lateral tilt, axial rotation and flexion/extension, lateral flexion and axial rotation respectively) were compared between players with and without lumbar spine abnormalities, genders, and serve types using a mixed-effects model. Exploratory data pertaining to the order and timing of key serve events was also collected. RESULTS: Males had significantly greater posterior pelvis tilt than females during the drive phase of both flat (M, - 7.1 ± 5°; F, 4 ± 5.5°) and kick serves (M, - 8.6 ± 5.1°; F, 2.1 ± 5.8°). Independent of serve type, males also impacted the ball ~ 15 cm further into the court than females, while all players contacted flat serves significantly further forward (~ 17 cm). There were no effects for abnormality in the magnitude of pelvis and trunk kinematics. The order and timing of key serve events, however, did tend to differ between those with and without lumbar spine abnormalities. Players with abnormalities entered peak front knee flexion and initiated pelvis rotation earlier than players without abnormalities. Lastly, the timing of pelvis rotation was highly variable among females though not males. CONCLUSION: Pelvis and ball toss kinematics vary with gender and serve type but not necessarily abnormality in the elite adolescent serve. There is evidence to suggest that the order and timing of key serve events might help to identify those at risk of lumbar spine abnormalities; however, further research is needed to investigate the statistical significance of the timing of these events.