Conflicting conclusions have been drawn regarding cochlear distortion products from cochlear microphonic (CM) studies [e.g., Dallos,The Auditory Periphery(Academic, New York, 1973)], and from neural studies [Kim and Molnar, inThe Nervous System, edited by Tower, (Raven, New York, 1975), and Kimet al., Scand. Audiol. (in press)]. The former studies conclude that there are no mechanically propagated distortion products in CM responses for sound pressure levels (SPLs) below 70 dB (re20 μN/m2). The latter studies however, conclude that the distortion products (f2‐f1and (2f1‐f2) are (1) generated in the cochlear region where the primary‐frequency responses are both large, and (2) propagate mechanically like single‐tone responses over the region apical to the generation region even at low SPLs (e.g., 34 dB). To help resolve this conflict, we measured CM responses in the chinchilla. We sought and found predominant distortion products (f2‐f1) and (2f1‐f2) by systematic variation of stimulus frequencies and SPLs with distortion frequency near the characteristic frequency (CF) of the recording site (turn two or three). We subsequently varied the distortion frequency around the CF, e.g., by fixingf1well above the CF and varyingf2. We have observed that the distortion amplitude in CM obtained with SPLs as low as 25 dB shows tuning characteristics similar to the single‐tone response. We conclude, in agreement with the neural studies, that our CM data reflect mechanically propagating distortion products. Possible explanations for the previous contrary conclusions are discussed. [Supported by NIH Grants NS07498, NS07057, NS00162, RR00396.]