Wrong-way driving is rare. That is the first thing to say, because context matters and because fear is not a useful safety tool.
It is also getting worse, and the trend has been consistent enough that it is no longer reasonable to treat it as noise.
The numbers
Research published by the AAA Foundation for Traffic Safety examined fatal wrong-way crashes on divided highways across the United States from 2014 through 2023. Over that decade there were 4,164 fatal wrong-way crashes, and those crashes killed 5,730 people.
The trend inside that decade is what matters most. The annual number of these crashes approximately doubled — from 278 in 2014 to 520 in 2023.
That growth outpaced the overall trend. Wrong-way crash fatalities rose faster than fatal crashes generally over the same period, meaning this is not simply a reflection of more driving or more crashes across the board. Something specific is getting worse.
Why "rare" and "serious" are both true
Wrong-way crashes are a small fraction of all traffic deaths. Any honest discussion has to start there.
But rarity is not the same as unimportance, for two reasons.
The first is severity. A wrong-way collision is typically head-on, at combined highway speed, with very little time for either driver to react. The physics leave far less room for survival than almost any other crash type. That is why a small number of events produces a disproportionate number of deaths.
The second is preventability. Many traffic deaths involve a complicated mix of factors that are difficult to address. Wrong-way crashes have an unusually clear structure: a vehicle enters a roadway in the wrong direction, usually at an identifiable point, usually under identifiable conditions. That structure is what makes prevention realistic.
Who and when
The AAA Foundation research identified groups at elevated risk. Older drivers, alcohol-impaired drivers, drivers without valid licenses, and drivers of older vehicles were all overrepresented.
It also identified circumstances associated with higher risk: rural areas, dawn and dusk, nighttime, drivers who lived far from where the crash occurred, and drivers operating vehicles registered to someone else.
Read together, those factors sketch a recognizable picture — impairment, unfamiliarity with the road, and reduced visibility, often in combination, often at night. That is not a random distribution. It is a set of conditions, and conditions can be targeted.
What this means for prevention
If wrong-way crashes were evenly distributed across drivers, roads, and hours, prevention would be nearly impossible. Because they cluster, it is not.
It means detection technology can be concentrated where entries actually happen. It means enforcement and awareness can focus on the hours that matter most. It means road design changes at high-risk interchanges have outsized value relative to their cost. And it means the warning that reaches a driver seconds earlier is worth a great deal, because seconds are what this crash type takes away.
Arizona has been one of the more active states on this, and we will look closely at what has been deployed here and what it has actually detected in a coming post.
The honest framing
We are not trying to convince anyone that wrong-way driving is the largest road safety problem in America. It is not.
We are saying that it is a problem that is measurably growing, that kills people in a particularly brutal way, and that has a clear enough structure to be worth attacking directly. That combination — rare, severe, and preventable — is exactly where focused effort pays off.