When an auditory stimulus is presented at a repetition ratefrof about 40/s, a robust steady‐state auditory evoked potential (AEP) is recordable [R. Galamboset al., Proc. Natl. Acad. Sci. (USA)78,2643–2647 (1981)], and has an approximate sinusoidal waveform with frequencyfr. A similar phenomenon is encountered when auditory evoked magnetic fields (AEMFs) are measured at the scalp with a SQUID magnetometer. We have determined thefrdependence of the steady‐state AEPs and AEMFs for the same human subjects. Using the assumption of a sandwich system (linear transformation‐zero memory nonlinear transformation‐linear transformation) response, we have also produced computer simulations of thefrdependence of steady‐state responses from measured transient‐response data. The simulations account fairly well for the steady‐state data. The stimulus was a 500‐Hz tone pip presented withfrvarying from 3‐50/s. The AEMF was recorded from a point around theT4location, and the AEP was recorded between electrodes atCZand the mastoid process of the test ear. Preliminary results for optimalfr's are 40‐45/s for the AEPs and 30‐35/s for the AEMFs. [Work supported by NINCDS, NSF, and the Purdue University Biomedical Support Fund.]