This study was part of a series of studies on variable-reflectance rearview mirrors. Previous work included laboratory studies of human visual performance, field collection of photometric data, and mathematical modeling of the visual benefits of variable-reflectance mirrors. We extended that work in this study by collecting photometric and human-performance data while subjects drove in actual traffic.Three mirror conditions were investigated: (1) fixed-reflectance mirrors in the center and driver-side positions, (2) a variable-reflectance mirror in the center with a fixed-reflectance mirror on the driver side, and (3) variable-reflectance mirrors in both positions. The fixed and variable reflectivities were produced by the same mirrors by overriding the circuitry that normally controlled reflectance in the variable mode.Results indicated that variable-reflectance mirrors provided a substantial reduction in discomfort glare without a measurable reduction in subjective ratings of rearward seeing ability. They did not cause major improvements in forward seeing (in agreement with previous laboratory and modeling results).The present study is inconclusive with respect to the benefits of a variable-reflectance driver-side mirror relative to a fixed-reflectance mirror with 50% reflectance. The reason for this is that the particular driver-side mirror used in this study became noticeably green in the low-reflectivity state, and thus low reflectivity was confounded with a color change. The effect of the driver-side mirror should be clarified by further research.