The strength-duration time constant (tau SD) is a property of nodal membrane and, while it depends on a number of factors, its measurement may shed light on axonal properties when taken in conjunction with measurements of axonal excitability. For example, tau SD increases with demyelination as the exposed membrane is enlarged by inclusion of paranodal and internodal membrane, it decreases with hyperpolarization and it increases with depolarization. The present study was undertaken in 20 normal volunteers to compare strength-duration curves for compound sensory and muscle action potentials, to determine the most appropriate curve fitting equation for the data, and to examine the reproducibility of the calculated time constant on different days, for potentials of different amplitude and at different sites along the nerve. Using a computerized threshold-tracking system, stimulus intensity was adjusted to produce an antidromic compound sensory action potential (CSAP) or an orthodromic muscle action potential of 30% of maximum. Stimulus duration was increased every minute in 20 microseconds steps from 20 microseconds to 1 ms. The time constant for compound sensory potentials (665 +/- 182 microsecond) was longer than that for compound EMG potentials (459 +/- 126 microseconds). Weiss's formula, which relates threshold charge to stimulus duration, provided an accurate fit for the experimental data, and the study validated that, using it, relatively few experimental measurements were required to calculate the time constant. In repeated studies on the same subject, time constants usually differed by < 400 microseconds for sensory axons and < 250 microseconds for motor axons. They were identical at different sites along the nerve and did not alter with the size of the compound action potential. These characteristics suggest that the determinations of strength-duration time constant could be suitable for clinical usage.