Musculoskeletal modelling plays a crucial role in understanding joint mechanics, particularly in applications such as surgical planning and implant design. As a common approach, these models are generally validated by assessing their ability to predict knee contact forces. However, such validation may not necessarily guarantee an accurate reconstruction of the complete joint biomechanics, where predicted kinematic patterns are often neglected, which is critical for understanding soft tissue loading and wear/interface conditions. In this study, we used a musculoskeletal model of the knee incorporating detailed representations of articular contact and soft tissue constraints to explore the relationship between the rigor of knee contact force validation and uncertainties in kinematic predictions. A Monte Carlo simulation with 1000 variations in muscle activation strategies was conducted, using a cost function that minimized the sum of squared muscle activations. The resulting outcomes of level walking and squatting simulations were then analysed.

Our findings indicate that simulations yielding appropriate knee contact force estimates do not necessarily guarantee precise predictions of joint kinematics. Specifically, extending the acceptable root mean square error range for knee contact force estimates by 15 % of body weight led to an increase in the uncertainty of kinematic outcomes, reaching approximately 8 mm in translations and 10° in joint rotations. Stricter force validation criteria may mitigate, but not eliminate, inaccuracies in kinematic predictions. Our results highlight the need for comprehensive validation that includes both kinetic and kinematic data to achieve robust modelling outcomes. This is especially critical in applications requiring precise joint mechanics, such as implant design and in silico wear prediction.