Autonomous vehicles are expected to influence daily travel significantly. Despite autonomous vehicles’ potential to enhance safety and to reduce congestion, energy consumption, and emissions, many studies suggest that the system-level effects will be minimal at low market penetration rates. Introducing reserved lanes for autonomous vehicles is one potential approach to address this limitation because these lanes increase autonomous vehicles’ density. However, preventing regular vehicles from using specific lanes can significantly increase congestion in other lanes. Accordingly, this study explored the potential effects of reserving one lane for autonomous vehicles on traffic flow dynamics and travel time reliability. A two-lane hypothetical segment with an on-ramp and a four-lane highway segment in Chicago, Illinois, was simulated under different market penetration rates of autonomous vehicles. Three strategies were evaluated: ( a) mandatory use of the reserved lane by autonomous vehicles, ( b) optional use of the reserved lane by autonomous vehicles, and (c) limiting autonomous vehicles to operate autonomously in the reserved lane. Policies based on combinations of these strategies were simulated. It was found that optional use of the reserved lane without any limitation on the type of operation could improve congestion and could reduce the scatter in a fundamental diagram. Throughput analysis showed the potential benefit of reserving a lane for autonomous vehicles at market penetration rates of more than 50% for the two-lane highway and 30% for the four-lane highway. Travel time reliability analysis revealed that the optional use of the reserved lane was also significantly beneficial.