Current understanding of how muscles coordinate walking in humans is derived from analyses of body motion, ground reaction force and EMG measurements. This is Part I of a two-part review that emphasizes how muscle-driven dynamics-based simulations assist in the understanding of individual muscle function in walking, especially the causal relationships between muscle force generation and walking kinematics and kinetics. Part I reviews the strengths and limitations of Newton–Euler inverse dynamics and dynamical simulations, including the ability of each to find the contributions of individual muscles to the acceleration/deceleration of the body segments. We caution against using the concept of biarticular muscles transferring power from one joint to another to infer muscle coordination principles because energy flow among segments, even the adjacent segments associated with the joints, cannot be inferred from computation of joint powers and segmental angular velocities alone. Rather, we encourage the use of dynamical simulations to perform muscle-induced segmental acceleration and power analyses. Such analyses have shown that the exchange of segmental energy caused by the forces or accelerations induced by a muscle can be fundamentally invariant to whether the muscle is shortening, lengthening, or neither. How simulation analyses lead to understanding the coordination of seated pedaling, rather than walking, is discussed in this first part because the dynamics of pedaling are much simpler, allowing important concepts to be revealed. We elucidate how energy produced by muscles is delivered to the crank through the synergistic action of other non-energy producing muscles;​ specifically, that a major function performed by a muscle arises from the instantaneous segmental accelerations and redistribution of segmental energy throughout the body caused by its force generation. Part II reviews how dynamical simulations provide insight into muscle coordination of walking.