Recordings were made of discharges in single fibers of the cat’s auditory nerve using single‐tone and two‐tone signals of varying frequency. The frequency of the single tone was varied stepwise from the fiber’s best frequency to about 1.7 times this value. One component of the two‐tone signal was varied in exactly the same manner while the frequency of the other was increased in such a way that the frequency of the cubic difference tone (CDT, 2f1−f2) generated within the auditory system was kept constant at the fiber’s best frequency. Responses in terms of discharge rate and period histograms were collected for both signals. The difference in spike rate between the response to the single tone and to the two tones could be ascribed to the CDT and was used to estimate the CDT level as a function of the signal level and frequency. It turned out that the CDT level relative to that of the primaries is about 15 dB smaller for frequencies beyond 10 kHz than for lower frequencies, and slighly decreasing for an increasing stimulus level. Period histograms of responses to CDTs of frequencies below 4 kHz made it possible to estimate CDT phase as a function of the frequencies and the level of the two generating tones. The intensity of the tones had no effect on the CDT phase. The varying frequency separation produced shifts in the CDT phase of several hundred degrees if its frequency was above 2.5 kHz. For lower frequencies the shifts were maller, and varied considerably between different preparations. The phase estimated from period histograms for acoustic tones and for CDTs of varying frequency turned out to change in an identical manner. This shows that the CDT must be subject to the same delays or phase shifts as the traveling wave in the cochlea produced by an acoustic tone.