1. Intra-axonal recording and electron microscopy were applied to intramuscular myelinated axons in lizards and rats to investigate factors that influence the amplitude and time course of the depolarizing after-potential. 2. Depolarizing after-potentials in lizard axons had larger peak amplitudes and longer half-decay times than those recorded in rat axons (mean values 10 mV, 35 ms in lizard; 3 mV, 11 ms in rat). These differences were not due to differences in temperature, resting potential or action potential amplitude or duration. 3. For a given axon diameter, the myelin sheath in lizard fibres was thinner and had fewer wraps than in rat fibres. There was no significant difference in myelin periodicity. Calculations suggest that the thinner myelin sheath accounts for < 30% of the difference between depolarizing after-potential amplitudes recorded in lizard and rat axons. 4. Consistent with a passive charging model for the depolarizing after-potential, the half-time of the passive voltage transient following intra-axonal injection of current was shorter in rat than in lizard axons. 5. Aminopyridines prolonged the falling phase of the action potential and increased the amplitude of the depolarizing after-potential in both types of axon. 6. During repetitive stimulation the depolarizing after-potentials following successive action potentials exhibited little or no summation. Axonal input conductance in the interspike interval increased during the train. 7. These findings suggest that the amplitude and time course of the depolarizing after-potential are influenced not only by the passive properties of the axon and myelin sheath, but also by persisting activation of axolemmal K+ channels following action potentials.