Two versions of a cascade add, attenuate, and delay circuit were used to generate iterated rippled noise (IRN) stimuli. IRN stimuli produce a repetition pitch whose pitch strength relative to the noise percept can be varied by changing the type of circuit, the attenuation, or the number of iterations in the circuit. Listeners were asked to use pitch strength to discriminate between various pairs of IRN stimuli which differed in the type of network used to generate the sounds, the number of iterations, and the attenuation in the network. A description based on an exponential function of the first peak of the autocorrelation function of IRN stimuli is consistent with the results. The discrimination data were well fit by a function based on the difference in the exponential functions of the first peak of the autocorrelation functions. A magnitude estimation experiment indicated that pitch strength was an exponential function of the height of the first peak in the autocorrelation function. These results suggest that the strength of the pitch of IRN stimuli is based on temporal processing as might be revealed by autocorrelation.