Background: Observing how individuals actively adapt to their environment may provide additional insights into traditional clinical tests. Rather than using tests that only identify joint mobility limitations, it seems relevant to use clinical motion tests that assess global biomechanical functions more generally and identify functional behaviours. Objectives: This study explores whether different functional kinematic behaviour patterns appear when executing a new complex motor task and whether those observations are consistent over multiple executions. Methods: Marker-based kinematic analyses of the lumbo-pelvic complex were conducted on 29 asymptomatic athletes during two active self-induced motion tests: the one-sided tilt test and a modified version of this test limiting the trunk axial rotation. Marker data served as an input for a full musculoskeletal model to compute the lumbar and lower limb joint angles. Latent class analysis and intraclass correlation coefficients were calculated to identify different classes of functional kinematic behaviour and assess the reliability between measurements. Results: The methodology allowed us to identify four distinctive classes of possible movement combinations based on these two functional tests: standard movement, low knee and lumbar engagement, high pelvis engagement and high lumbar flexion. All ICCs for the lumbo-pelvic complex degrees of freedom were higher than 0.6, suggesting a moderate to good reliability for the overall test. Conclusion: It remains unknown whether the observed reproducible patterns emerging from the motion test relate to motivation and prior experiences. Further exploration is required to investigate whether these behaviours can be correlated to empirical clinical observations, past experiences, and future vulnerabilities for musculoskeletal conditions