The amplitudes of inaudible subjective signals are estimated using the monaural tone‐on‐tone masking procedure. These inferences are based upon three major assumptions: phase‐locked generation of subjective signals, vector summation between a subjective and acoustic tone, and the independence of the distortion from the sensitivity processes. Accordingly, any shift in sensitivity without an accompanying change in distortion levels should produce quantifiably predictable variations in the measured phase‐amplitude effects. Since the assumptions would most likely be violated at low frequencies, this prediction has been tested with a 250‐Hz masker, and comparisons of relatively long‐ versus short‐duration thresholds for a 500‐Hz maskee (probe). Observed decreases in phase‐magnitude effects with increasing losses in sensitivity follow predictions. These findings: (1) support the validity of the estimation procedure among subjects with normal hearing, (2) suggest a control condition important in estimating subjective tones among abnormal‐hearing subjects, and (3) have implications for inferring the origins of the monaural phase effects widely observed with inharmonic and harmonic probe tones of relatively high intensity. [Work supported by NIH and the Department of Otorhinolaryngology, University of Michigan Medical School.]