In this experiment 12 experienced truck drivers drove a fixed-base driving simulator for three 8-h sessions under simulated nighttime driving conditions. Sessions included (a) no glare, (b) intermittent glare presented in the exterior rearview mirrors to simulate following vehicles, and (c) intermittent glare with electrochromic glare reduction. The driving task combined vehicle control on straight and curved road segments with detection of pedestrians appearing alongside the road and targets appearing in the rearview mirrors. The presence of glare slowed detection of pedestrians and, to a lesser extent, slowed the detection of targets appearing in mirrors. Glare was also associated with increased lane position variability, reduced speed on curves, and, most consistently, increased steering variability. We found only meager evidence that electrochromic glare reduction improved target detection performance and no evidence that glare reduction improved vehicle control, despite the fact that participants consistently voiced positive preferences for glare reduction. The results will aid decision making that requires incorporation of the benefits of electrochromic glare-reducing mirrors.