Electrical stimulation of neural responses is used both scientifically in brain mapping studies and in many clinical applications such as cochlear, vestibular, and retinal implants. Due to safety considerations, stimulation of the nervous system is restricted to short biphasic pulses. Despite decades of research and development, neural implants are far from optimal in their ability to restore function and lead to varying improvements in patients. In this study, we provide an explanation for how pulsatile stimulation affects individual neurons and therefore leads to variability in restoration of neural responses. The explanation is grounded in the physiological response of channels in the axon and represented with mathematical rules that predict firing rate as a function of pulse rate, pulse amplitude, and spontaneous activity. We validate these rules by showing that they predict recorded vestibular afferent responses in macaques and discuss their implications for designing clinical stimulation paradigms and electrical stimulation-based experiments.