The dynamic properties of excitation and two-tone inhibition in the cochlear nucleus were studied from extracellularly recorded unit responses to two simultaneously presented tones. One tone was presented at the unit's characteristic frequency, CF, the other at the unit's best inhibitory frequency, BIF. One or both of the tones were amplitude-modulated with pseudorandom noise. The system under study is in general nonlinear, but can be considered to function as a linear system for small changes in sound intensity around a certain operating point. The dynamic properties are likely to be different at different operating points. A suitable method for the study of dynamic properties of such a system employs tones that are amplitude-modulated with pseudorandom noise. In the present study, the dynamic properties were assessed by cross-correlating the unit discharge rate with the modulation. This was accomplished by computing the cross-covariance function between a period of noise and a period histogram of the discharges, the histogram being locked to the periodicity of the pseudorandom noise. In this way, it has been shown in previous works (Moller, 1973, 1974b), that the cross-covariance function is a valid approximation of the system's impulse response function at a certain sound intensity, provided the modulation is kept at a low value. In the present study the computed cross-covariance function is thus an approximation of the change in discharge rate of the cochlear nucleus units in response to a brief increase in stimulus intensity. As the response of the system under the given circumstances is approximately that of a linear system, the integrated cross-covariance is an approximation of the system's step response function, i.e the change in discharge rate that resulte from a hypothetical step increase in stimulus intensity. The results of the present study can be summarized as follows: 1. The impulse and step response functions computed from the responses to the modulated inhibitory tone of the great majority of units from which recording was made were found to be virtual mirror images of those obtained when the excitatory tone was modulated, the inhibitory response being somewhat smaller in amplitude than the excitatory. 2. When both tones were modulated simultaneously, the step response function was approximately the algebraic sum of the two responses obtained when the tones were modulated singly, further indicating that the system functions as a linear system when the stimulus amplitude is varied slightly around a certain operating point. 3. The shape of the cross-covariance functions is similar for all three stimulus situations, but varies with stimulus intensity and is different in different units. 4. The implication of the results is that the inhibition studied may either originate from the inhibition (suppression) seen in primary fibers or it may be the result of a true neural inhibition in the cochlear nucleus that occurs without any interneurons.