Human nerve fibers exhibit a distinct pattern of threshold fluctuation following a single action potential known as the recovery cycle. We developed geometrically and electrically accurate models of mammalian motor nerve fibers to gain insight into the biophysical mechanisms that underlie the changes in axonal excitability and regulate the recovery cycle. The models developed in this study incorporated a double cable structure, with explicit representation of the nodes of Ranvier, paranodal, and internodal sections of the axon as well as a finite impedance myelin sheath. These models were able to reproduce a wide range of experimental data on the excitation properties of mammalian myelinated nerve fibers. The combination of an accurate representation of the ion channels at the node (based on experimental studies of human, cat, and rat) and matching the geometry of the paranode, internode, and myelin to measured morphology (necessitating the double cable representation) were needed to match the model behavior to the experimental data. Following an action potential, the models generated both depolarizing (DAP) and hyperpolarizing (AHP) afterpotentials. The model results support the hypothesis that both active (persistent Na(+) channel activation) and passive (discharging of the internodal axolemma through the paranodal seal) mechanisms contributed to the DAP, while the AHP was generated solely through active (slow K(+) channel activation) mechanisms. The recovery cycle of the fiber was dependent on the DAP and AHP, as well as the time constant of activation and inactivation of the fast Na(+) conductance. We propose that experimentally documented differences in the action potential shape, strength-duration relationship, and the recovery cycle of motor and sensory nerve fibers can be attributed to kinetic differences in their nodal Na(+) conductances.