The effect of a preceding brief interruption in the illumination of a small numeral eight on the auditory startle blink was investigated in human subjects. In Exp. 1 (n = 9) trials were given in which a 20-msec. interruption of the visual stimulus preceded elicitation of the startle blink by a loud tone at intervals from 80 to 200 msec. Compared to trials on which the digit was continuously lit but the reflex was elicited, the blink was augmented at 80 msec. and depressed at longer intervals. Exp. 2 (n = 24) showed that the depressive effect peaked at about 200 msec. and did not depend upon requiring subjects to make judgements concerning the stimuli. In Exp. 3 (n = 17) dark intervals ranged from 1.25 msec. to 40 msec. and preceded the loud tone by 200 msec. Reflex effects paralleled subjects' reports: dark intervals of 10 msec. or greater reliably inhibited the reflex and subjects reported seeing the dark interval at least 50% of the time at intervals of 8 msec. or greater, suggesting that the reflex method may be useful as an alternative objective technique for measuring visual temporal resolution.