Level and phase of the 2f1−f2 combination tone were measured in five human subjects using two methods: Psychoacoustic cancellation and direct cancellation of the distortion-product emission (DPE) measured in the ear canal. A probe assembly, composed of two earphones and a microphone, was placed in the subject's ear canal for the delivery of stimulus tones, f1 and f2, and the 2f1−2f2 cancellation tone and for measurement of DPEs. Frequency of f1 was 1620 Hz with f2 either 1800, 1851, 1944, or 2052 Hz. Level of f1 or f2 was varied in 5-dB steps from 70–30 dB SPL for each frequency pair. Data from these subjects were compared with similar data measured using the hardware model. DPE cancellation levels were 40–50 dB below comparable psychoacoustical cancellation levels for both data sets, however absolute differences varied by frequency separation and level of the primaries and by method. Dependencies of the level and phase of 2f1−f2 on the frequency separation and level of the primaries were similar for all three methods.