Thresholds for the detection of a tone added in-phase to the carrier of a fully modulated SAM tone were measured. In some conditions the signal was added to a single SAM tone, and in other conditions the signal was added to the sum of three or more SAM tones. Level equalization ensured that the addition of the tonal signal did not lead to increases in energy. When multiple SAM tone maskers were used, a small number of reproducible maskers were tested, each masker being composed of SAM tones with a variety of relative modulator phases. The maskers were eitherfixedacross intervals and trials,rovedacross trials but fixed across intervals, orrandomlychosen across both intervals and trials. The frequency separation between the signal-centered and off-frequency SAM tones was also varied. For small signal-centered/off-frequency SAM tone frequency separations, a separation ratio of 1.3, thresholds in thefixedcondition depend on the relative modulator phases, and a simple mixture model reasonably predicted thresholds in therovingcondition based on thresholds in thefixedcondition for two of the three observers. For signal-centered/off-frequency SAM tone frequency separation factor of 1.68, effects of relative modulator phases were not obtained. Thresholds in thetarget-alonecondition were generally superior to thresholds measured with the comodulated masker. Comodulated thresholds were better thantarget-alonethresholds only when level equalization was not used, and so the addition of the signal led to increases in level.