Pairs of noisebursts (75 dB SPL, 150 msec apart) were presented to subjects who were instructed to judge the duration of one of them (the target) and ignore the other (the conditioner). The two noisebursts could be presented to the same ear (monaural) or to opposite ears (dichotic), and the conditioner could precede or follow the target (forward and backward paradigm, respectively). Subjects estimated target duration by changing the duration of a third noiseburst (also 75 dB SPL) presented 1500 msec after the pair. Five conditioner (5, 10, 20, 40, and 80 msec) and two target durations (20 and 25 msec) were used. In the forward monaural paradigm the average subject selected bursts of 17 and 26 msec as matching the 20 msec target when the conditioner was set at 5 and 80 msec, respectively. The matches showed smaller and more variable shifts in the monaural backward, dichotic forward, and dichotic backward paradigms. These perceptual duration shifts resemble a variety of other auditory illusions, such as loudness enhancement and pitch shifts (R. Elmasian and R. Galambos, J. Acoust. Soc. Am.58, 229–234 (1975);58S83 (1975)] and may be related to them by a common mechanism.