The configuration of a musculoskeletal (MS) system is closely related to the individual motions of the human body. Many researches have been focused on evaluating the associations between the MS configuration and the individual motion using patient-specific MS models, but it still remains a challenging issue to accurately predict the motion by differed configurations of the MS system. The main objective of this paper is to predict the changes of a patient-specific gait by altering the geometric parameters of the hip joint using function-based morphing method (FBM). FBM is suitable for motion analysis since this method provide a robust way to morph a MS model while preserving the biomechanical functions of the bones. Computed-muscle control technique is used to calculate the muscle excitations to reproduce the targeted motion within a digital MS model without the motion-captured data. We applied this approach to a patient who has an abnormal gait pattern. Results showed that the femoral neck length and the angle significantly affect to the motion especially for the hip abduction angle during gait, and that this approach is suitable for gait prediction.