Two experiments were conducted in the laboratory to evaluate potential benefits of different retroreflective markings for nighttime pedestrian visibility. Video recordings of a jogger wearing four different markings were made from a vehicle in four different road environments. Subjects viewed composite tapes that included each of the 16 jogger marking/road environment combinations as well as travel with no targets. The task was to step on a pedal immediately upon seeing a jogger, which had no effect on the flow of the video playback. The time between depression of the pedal and the point of "impact" was the major dependent variable. Experiment I showed that performance was better for all retroreflective markings than for the dark control and that it was better with markings of the limbs than of the torso. Experiment 2, which included a secondary video tracking task, showed that performance was better for markings that incorporate biological motion than for a vest or arbitrarily positioned stripes on the limbs. Questionnaire data indicated that 85% of the subjects judged the biological motion markings to be "easiest to see." Also, subjects reported more conservative estimates of nighttime visibility and greater willingness to take personal precautions at night after participating in the experiment.