Around the world, ungulates must cross a matrix of highways to reach the resources they need. Wildlife-vehicle collisions (WVCs) on these roads are a significant source of wildlife mortalities, and traffic hinders wildlife movements. Wildlife crossing structures can reduce WVCs by ~90%; however, these structures are expensive and sometimes impractical, prompting transportation managers to seek alternatives. Among the general public, there is support for reducing speed limits in WVC hotspots. However, almost no research exists on the effectiveness of reducing speed limits as a way to reduce WVCs, and studies of driver behavior suggest that reducing the posted speed limit may not slow drivers down. In order to inform this debate and future management practices, we worked with the Wyoming Department of Transportation (WYDOT) to assess the effectiveness of reduced nighttime speed limits in six deer-vehicle collision hotspots in western Wyoming. In each hotspot, WYDOT reduced the posted speed limit from 70 mph to 55 mph at night during peak WVC seasons. Speed limit signs indicated a daytime and nighttime speed and used flashers to alert drivers when the night speed limit was in effect. We used a BACI design over two winters; in Y1, speed limit was maintained at 70 mph in all locations, and in Y2 the reduced speed limit treatment was implemented in half of each hotspot (hereafter "stretch"). We measured the effects of these speed limit reductions on vehicle speeds and traffic dynamics using radar recorders; deer road-crossing behavior using infrared video cameras; and deer-vehicle collision rates using carcass counts and reported collisions. Vehicle speed patterns were remarkably consistent across all six sites. In response to a night speed limit that was 15 mph lower than daytime speed limit, drivers consistently reduced their speed by 3-5 mph (significant year by stretch interactions in ANOVA analyses of the BACI design). Cargo vehicles traveled at slightly slower speeds than passenger vehicles, and drivers of both reduced their speeds similarly in response to the posted speed limit reduction. Although drivers were clearly and consistently responsive to the reduced speed limit, there was no evidence that this resulted in lower rates of wildlife-vehicle collisions. We used chi-squared tests to compare collision and carcass counts in paired stretches in Y1 and Y2. Neither collision data nor carcass data showed any reductions in response to the speed limit changes in Y2 (collision: chi-squared = 2.21, p=0.14; carcass: chi-squared = 0.03, p=0.86). Based on these results, we conclude that reduced night speed limits on high-speed rural two-lane highways does not cause drivers to slow enough to reduce their impacts on wildlife. This study is unique in scope and will be important in determining whether reduced speed limits are adopted more widely across the United States.