The method of best beats has been employed to estimate the intensities of aural harmonics and of combination tones. It has been generally assumed that the listener hears best beats when the exploring tone produces in the cochlea a disturbance that is equal in magnitude to that of the aural harmonic or of the combination tone being measured. However, as the experiments to be reported will show, when the tone to be measured is near the absolute threshold or is partially masked, the most prominent beats will be heard when the intensity of the exploring tone exceeds that of the unknown tone. Consequently, since aural harmonics and combination tones are partially masked, their intensities will be overestimated when the method of best beats is used. The present paper concerns the magnitude of this error of overestimation.A procedure is presented by which a better estimate of the intensity of an aural harmonic or of a combination tone may be made. When the tone being measured by the method of best beats is well above threshold, the listener hears beats over a wide range of intensities of the exploring tone, and the error of over‐estimation is small. However, when the tone being measured is near threshold, there is a small range of audible beats, and the error of overestimation is of considerable magnitude. It is therefore possible to correct for this error by determining the range of intensities of the exploring tone over which beats are audible as well as the intensity of the exploring tone required for best beats. Application of this procedure to the measurement of a second aural harmonic is illustrated.An explanation is given of the fact that the intensity of a tone near its absolute threshold will be overestimated by use of the method of best beats. This explanation is formulated in terms of the relation of the minimum and the maximum of the envelope of the beating complex to the listener's threshold for the tones that beat.