In a previous experiment, Martin, Siegel, Johnson, and Haroldson (1984) found that stutterers reduced their stuttering under amplified sidetone, but only if the amplified sidetone condition had been preceded by a condition of speaking in noise. The authors speculated that when stutterers were exposed initially to loud noise and its attendant reduction in stuttering, they became "sensitized" to reduced stuttering in any subsequent condition where their auditory feedback was modified. The current experiment tested that hypothesis with 24 adult stutterers divided evenly into three groups. One group of stutterers spoke with no auditory stimulation (quiet), then while receiving 100 dB SPL white noise, and then while receiving amplified sidetone (0, +10, +20 dB SPL). A second group spoke in quiet, then while receiving amplified sidetone, and then while receiving amplified sidetone again. A third group spoke in quiet, then while receiving rhythmic stimulation, and then while receiving amplified sidetone. Relative to vocal intensity in quiet, stutterers spoke with increased vocal intensity during noise, with decreased vocal intensity during amplified sidetone, and with no significant change in vocal intensity during rhythmic stimulation. Relative to stuttering frequency in quiet, subjects spoke with decreased stuttering frequency during loud noise and during amplified sidetone, but only when the amplified sidetone was preceded by the loud noise condition. These results are discussed in terms of the "sensitization" hypothesis.