In this study, the effects of a hypothetical autonomous vehicle (AV)-exclusive roadway were estimated through a step-by-step approach using both microscopic and macroscopic simulations. First, the AV-exclusive roadway was classified into four types—entry lanes, mainlines, merging lanes, and intersections—and the C, α, and β values of the Bureau of Public Roads (BPR) function were estimated for each type through a microscopic simulation. These estimated values were then applied to a 3×3 (20 km) network, and a macroscopic simulation was conducted to compare the effectiveness of AVs and conventional vehicles (CVs) in terms of traffic volume and travel time.The analysis showed that for the same travel time, the traffic volume increased by more than 12% with AVs compared to that with CVs. Conversely, for the same traffic volume, the total travel time decreased by 11% for AVs. The estimated capacity of the AV-exclusive roadway, similar to the U-Smartway with a size of 3×3 (20 km), was approximately 400,000 vehicles, which was more than 140% higher than that of CVs. Assuming that each AV carries five passengers, up to two million people can be transported per day, indicating a significant potential benefit. However, these results were based on theoretical analyses using hypothetical networks under various assumptions. Future studies should incorporate more realistic conditions to further refine these estimations.