Walking is a primary therapeutic goal since it significantly improves physical health and general well-being, especially in children with motor disabilities. Nevertheless, the majority of children with cerebral palsy (CP), the most common mobility disability in childhood, lose the ability to walk. Early research on robotic gait trainers has produced encouraging results. However, the scope of these clinic-based solutions is restricted to brief programmes that are too short to sustain enhanced function in a lifetime handicap like cerebral palsy. CNS makes continuous adjustments through the musculoskeletal system to generate suitable joint stability by modulating the multi-joint stiffness through subtle adjustments in limb posture and muscle contraction level. Understanding lower limb musculoskeletal stiffness variations in CP-affected individuals can be used to design a more efficient design and control for the robotic device. This work uses a cable-driven serial chain model for the lower limb musculoskeletal system during the swing phase of walking to compare the lower limb multi-joint stiffness variations in healthy and CP individuals. The number of cables and routing are adapted from the available lower limb musculoskeletal structure. A multi-joint stiffness matrix, hip and knee joint, is formulated for the serial chain system considering different numbers of muscles. Inferences on the joint stiffness variations observed at hip and knee joints of the CP individuals when compared with healthy musculoskeletal systems were presented.