The most prominent auditory evoked field component occurs at about 100 ms latency and is termed the M100. We recorded M100 data from 20 subjects, in both hemispheres. We modeled the generators with a single equivalent current dipole in a 10 ms sliding window from 0 to 245 ms post-stimulus. A residual error curve was plotted, and a search for local minima identified two latencies at about 75 and 100 ms. In the left hemisphere, the early generator was about 6 mm above the later source; in the right hemisphere the early source was about 3 mm above the later, and 11 mm posterior. The M100 is a compound source, and the model may provide additional information in cases with reported laterality differences.