Detectability was measured as a function of signal duration for a dichotic condition in which there was only an interaural difference in level and for one in which there was only an interaural difference in time. For four subjects, performance improved as duration increased, and it did so in essentially the same way for the two dichotic conditions; however, one other subject was unusual in that he showed little improvement in performance as duration was increased when the sole interaural cue was a time difference. Masker and signal: the same band of noise, 50 Hz wide centered at 500 Hz. Method: single-interval forced choice.