Previous research has demonstrated that both emotional valence and arousal can influence the subjective experience of time. The current research extends this work by (1) identifying how quickly this emotional modulation of time perception can occur and (2) examining whether valence and arousal have different effects at different stages of perception. These questions were addressed using a temporal bisection task. In each block of this task, participants are trained to distinguish between two different exposure durations. Participants are then shown stimuli presented at a number of durations that fall between the two learned times, and are asked to indicate whether the test stimulus was closer in duration to the shorter or longer learned item. In the current study, participants completed blocks of trials in which the durations were “Short” (100–300 ms) or “Long” (400–1600 ms). Stimuli consisted of neutral photographs as well as four categories of emotional images: high-arousal negative, high-arousal positive, low-arousal negative, and low-arousal positive. In Short blocks, arousing and nonarousing negative images were judged to have been shown for shorter durations than they actually were (i.e., the duration was underestimated); this effect occurred at durations as brief as 133 ms. In Long blocks, the display time for highly arousing negative items was overestimated, whereas durations were underestimated for highly arousing positive items and less arousing negative items. These data suggest that arousal and valence have different effects at different stages of perception, possibly due to the different neural structures involved at each stage of the emotional modulation of time perception.