From near-parity in 2015 to a large urban lead in 2023

In 2015, rural and urban road fatalities were almost equal: 17,715 rural deaths versus 17,573 urban deaths out of a national total of 35,484. That near-parity was itself striking, since rural roads carry a substantially smaller share of total vehicle-miles traveled than urban roads. By 2023, the picture had shifted dramatically: 16,656 rural deaths against 23,921 urban deaths, with the national total reaching 40,901. Urban fatalities grew by approximately 36% over the eight-year window while rural fatalities fell by roughly 6%. In absolute terms, rural roads now account for approximately 41% of all traffic deaths in 2023, down from about 49.9% in 2015.

This crossing of the lines, urban deaths exceeding rural deaths in raw count, reversed a decades-long pattern. For most of the post-1970 period, rural roads produced the majority of traffic fatalities in the United States, driven by higher speeds, longer stopping distances, more severe single-vehicle run-off-road crashes, and greater distances to trauma centers. The 2015-2023 data shows urban deaths growing faster than rural ones during a period of rapid increases in urban vehicle-miles traveled, changing street design (more pedestrians and cyclists sharing arterial roads), and rising impaired-driving rates in urban cores.

What drove the urban surge

Several structural and behavioral changes contributed to the rise in urban fatalities between 2015 and 2023. Urban vehicle-miles traveled increased as ride-hailing platforms expanded, last-mile delivery fleets grew, and post-pandemic commuting patterns changed. More vehicles on urban streets at more hours means more exposure, and exposure is the single largest predictor of absolute fatality counts. The FARS data does not directly attribute this growth to any single cause, but the growth rate in urban deaths substantially exceeded the growth in urban VMT during this period, suggesting that per-mile urban risk also rose, not just absolute exposure.

Pedestrian fatalities, discussed in detail in a companion analysis, were a major driver of urban toll increases. The largest concentrations of pedestrian deaths occur on arterial roads in urban and suburban areas, which are frequently characterized by high speeds, long crossing distances, and limited pedestrian refuge. As America's vehicle fleet shifted toward larger SUVs and pickup trucks, the lethality of pedestrian strikes increased, independent of crash frequency. A pedestrian struck by a high-hood vehicle at 35 mph is far more likely to die than one struck by a low-hood sedan at the same speed, per studies from the Insurance Institute for Highway Safety and NHTSA's own pedestrian-protection research. This trend affected urban deaths disproportionately because urban areas are where pedestrian exposure is highest.

The paradox: rural roads are still far deadlier per mile

The headline shift from rural-dominated to urban-dominated absolute counts should not obscure the fundamental paradox: rural roads remain significantly more dangerous per mile driven. Rural roads carry roughly 30% of total vehicle-miles in the United States but produced approximately 41% of all traffic deaths in 2023. The implied per-mile fatality rate for rural roads is therefore roughly twice the national average. For urban roads, which carry around 70% of VMT but now account for roughly 58% of deaths, the per-mile rate is below the national average. Measured by the yardstick that matters most for road engineering and safety policy, rural roads remain the more dangerous environment.

Four interconnected factors explain persistent rural road danger, even as absolute counts have declined. First, speed: two-lane undivided rural roads often carry traffic at 55 to 65 mph, and high-speed crashes produce far more severe outcomes than lower-speed urban crashes. A side-impact collision at 60 mph on a rural road is categorically different from a side-impact at 25 mph in a city. Second, road geometry: rural roads frequently lack median barriers, rumble strips, clear zones, and guardrails that modern highway design standards require on high-speed facilities. Many rural routes were built decades before current safety standards and have not been retroactively upgraded. Third, distance from trauma care: survival after a severe crash depends critically on time to definitive surgical care. Rural crash victims average significantly longer EMS response and transport times than urban victims, and this delays-in-care factor is documented in NHTSA and HRSA research as independently contributing to rural fatality rates even at comparable crash severity. Fourth, seatbelt use rates: rural Americans consistently report lower seatbelt use rates in NHTSA observational surveys, and the protection effect of a seatbelt at 60 mph is enormous. The interaction between high speed, non-use of restraints, and distance from hospitals compounds the per-mile lethality of rural crashes.

Year-by-year context, 2015-2023

The trajectory was not simply a smooth urban climb. In 2020, both rural and urban deaths rose sharply as pandemic-era driving featured faster speeds on emptied roads and increased impaired driving during a period of social disruption. Urban deaths in 2020 jumped to 22,513 from 19,946 the prior year, a jump driven in part by increased speeds on emptied urban arterials and reduced law-enforcement patrol levels. Rural deaths in 2020 reached 16,340, roughly flat relative to 2019's 16,288, suggesting that the pandemic-era speed effect was concentrated in urban areas where speed changes were most dramatic on normally congested roads.

The 2021 spike to 43,230 total deaths was the largest single-year national toll since 2005. Urban deaths in 2021 reached 25,749, reflecting the full-year effect of behavioral changes that began in 2020. The subsequent partial recovery in 2022 and 2023 reduced the national total to 40,901 by 2023, primarily through declines in urban counts, with rural deaths holding relatively steady around the 16,600-16,700 range.

What the data means for policy

The rural-urban divergence between 2015 and 2023 argues for different policy emphases in different contexts. Urban road safety in this period was largely a pedestrian and speed-management problem, concentrated on arterial roads in built environments. The most effective interventions in urban settings are those that lower vehicle speeds on pedestrian corridors: road diet conversions, raised crosswalks, leading pedestrian intervals at traffic signals, and daylighting of intersections to improve sight lines. These have documented fatality-reduction effects in FHWA evaluation studies. Rural road safety, where per-mile danger remains elevated, is more effectively addressed through infrastructure hardening: adding rumble strips on two-lane roads, installing cable median barriers where right-of-way permits, improving delineation and curve-advisory signage, and improving trauma-center access through telemedicine and helicopter landing zones. Neither approach works everywhere, but the direction the data points is clear: urban roads have become the larger absolute problem while rural roads remain the more dangerous per-mile environment.