This study, part of a project on human factors in aircraft instrument lighting, has undertaken to determine how visual performance at low photopic brightness levels is affected by the brightness of an immediately preceding visual task. Two visual tasks were employed. In one, called the near task, subjects were required to read banks of photographic reproductions of instrument dials under instructions stressing speed and accuracy. The viewing distance was 28 inches, and three task brightnesses were employed: 2.9, 0.083, and 0.005 foot-lamberts. In the other, called the far task, subjects "read" banks of Landolt rings with speed and accuracy instructions. These were viewed periscopically at a distance of 18 feet, and five task brightnesses were used: 6.0, 0.076, 0.01, 0.007, and 0.0035 foot-lamberts. Subjects were high school and college students with excellent visual abilities. In Experiment I, subject (N=15) were visually adapted to the brightness level of the near task, performed the near task, then immediately were given the far task. All combinations of near and far brightnesses were used in a balanced experimental design. Experiment II was similar in all respects except that the subjects (N=12) first performed the far task then the near task. A third experiment was also carried out as a corroborative check on Experiment II.