Some of the data obtained by Hagiwara on the interval fluctuations of the sensory nerve impulse have been reanalysed with a modified model. The modification consisted essentially of the introduction of a factor characterizing the rate of fluctuation of a random process that is thought to be super-imposed on the threshold of the impulse triggering mechanism. The mean rate of change in amplitude of the step-wise defined random process was found to be in the order of a few hundred per second. Some of the implications of the model: the time course of the random function immediately after an impulse, the large value of the noise intensity relative to the rheobase and, as a consequence, the predicted existence of spontaneous discharge, were discussed.