Frequency-following responses, with latencies circa 6 msec, were recorded from five normal-hearing human subjects to brief 500 c/sec tone bursts presented monaurally. The frequency-following responses appear as peaks occurring at 2 msec intervals superimposed on a slow wave (pedestal-like) component. Comparisons were made between the frequency-following responses evoked by binaural and monaural stimuli. The results show that the binaural responses may be interpreted as the sum of two monaural responses. It is concluded, therefore, that there are two independent populations of neurons, each capable of generating a frequency-following response is not a microphonic-like response but rather that the individual waves in the frequency-following response are evoked by the collective activity of phase-locked single units. Finally, on the basis of the distinctness of the individual waves in the frequency-following response, it is concluded that the neural generators of the response must be spatially compact.