Preventing drunk driving requires a combination of personal choices, community strategies, and systemic policies. In 2022, 13,524 people were killed in alcohol-impaired driving crashes in the United States, accounting for 32% of all traffic deaths. That number has remained stubbornly high for years, but the strategies that work to bring it down are well established. They range from things you can do tonight to technology that could reshape driving safety within a decade.
Plan a Sober Ride Before You Drink
The simplest prevention happens before anyone picks up a glass. Designating a sober driver, pre-booking a rideshare, or choosing to stay where you’re drinking eliminates the decision entirely. What makes this effective isn’t willpower in the moment. It’s removing the choice while you’re still thinking clearly. Once alcohol impairs judgment, people routinely overestimate their ability to drive safely.
Rideshare services have measurably changed the landscape. When Uber launched in Seattle, Mothers Against Drunk Driving found a statistically significant 10% reduction in DUI arrests. A broader retrospective study found that after ridesharing entered new markets, the average proportion of alcohol-related crashes dropped from 39% to 29%, and the annual average of fatal alcohol-related crashes fell from 11.6 to 5. The effect isn’t universal, and some studies in larger metro areas have found no clear association, but in many communities the availability of a cheap, easy ride home has made a real difference.
How to Stop Someone From Driving Drunk
Knowing you should intervene and knowing how are two different things. Research on bystander intervention shows that the most successful approaches are practical rather than confrontational. In one study of people who actually stopped someone from driving drunk, the most common method (37%) was offering specific suggestions for getting home: calling a cab, offering your couch, walking together. Nearly as many people (36%) simply asked permission to drive instead. Only 6% demanded the person not drive.
The pattern is clear: giving someone a concrete alternative works better than telling them what not to do. Taking someone’s keys, telling them directly they’re too drunk, and asking them to stay longer all showed up in the research as strategies people attempt, but the ones framed around solutions rather than confrontation had more success. Public health messaging confirms this. Campaigns that model specific behaviors (“here’s what you say, here’s what you do”) improve people’s confidence and intention to intervene far more than vague slogans like “friends don’t let friends drive drunk.”
Two factors strongly predicted whether someone actually stepped in: feeling a moral or social obligation, and having consulted with other people at the gathering first. If you’re worried about a friend, talk to others at the party. Building a small consensus makes it easier to act and harder for the impaired person to brush off.
Ignition Interlock Devices
Ignition interlocks require a driver to blow into a breathalyzer before the car will start. They’re typically court-ordered for people convicted of impaired driving, and they work. Across six major studies, five found interlocks effective at reducing repeat offenses while installed. Drivers with interlocks were 15% to 69% less likely to be re-arrested for impaired driving compared to those who simply had their licenses suspended. The strongest evidence, from a randomized controlled trial, showed a 65% reduction in re-arrests.
The catch is that the benefit disappears once the device is removed. This is a limitation, but it also means that longer interlock periods provide longer protection. Many states now require interlocks even for first-time offenders, which expands the window of prevented incidents.
Sobriety Checkpoints and Enforcement
Sobriety checkpoints, where police stop drivers at predetermined locations to check for impairment, reduce alcohol-involved crashes consistently. The effect isn’t just about catching drunk drivers in the act. It’s about awareness. Studies from the CDC found that the largest increases in public awareness of checkpoints correlated with the largest decreases in alcohol-involved crash fatalities. States with higher enforcement levels and more publicity about their checkpoint programs saw clear reductions in deaths, while states with lower enforcement and less publicity did not.
The deterrent effect also lingers. An analysis of data from Los Angeles found that alcohol-involved crashes decreased for about a week following individual checkpoints. A similar study in Brisbane, Australia, confirmed the same pattern regardless of whether checkpoints were large or small, or lasted three hours or eight. States that conducted sobriety checkpoints had significantly lower rates of self-reported alcohol-impaired driving overall, even after controlling for binge drinking rates.
Lower Legal BAC Limits
Most of the United States sets the legal blood alcohol concentration limit at 0.08, but the global trend has moved well past that. Over 150 countries, including most of Europe, Asia, and South America, set their limits at 0.05 or lower. More than 6.5 billion people worldwide already live under these stricter limits. Eleven Canadian provinces and ten Mexican states have adopted limits between 0.04 and 0.05.
Utah became the first U.S. state to adopt a 0.05 limit in 2018 and saw a significant drop in impaired driving crashes afterward. The logic is straightforward: impairment begins well before 0.08, and a lower legal threshold discourages people from driving after even moderate drinking. At 0.05, most adults are noticeably impaired in reaction time, steering, and divided attention.
Zero Tolerance Laws for Young Drivers
Every U.S. state has a zero tolerance law making it illegal for drivers under 21 to operate a vehicle with any measurable alcohol in their system. These laws save lives at a scale that’s easy to quantify. Combined with underage purchase and possession laws, zero tolerance policies save an estimated 732 lives per year. The zero tolerance component alone accounts for roughly 159 of those lives annually, based on a 5% reduction in the ratio of drinking to nondrinking young drivers involved in fatal crashes.
Additional “use and lose” laws, which revoke a young person’s license for any alcohol violation, contributed another 5% decrease. The cumulative effect of all four major youth-targeted laws is an estimated 864 lives saved each year.
Training for Bartenders and Servers
Responsible beverage service training teaches bartenders and servers to recognize signs of intoxication and refuse service. When it works, the results extend beyond the bar. States with mandatory server training laws have seen lower rates of motor vehicle crashes involving high blood alcohol levels and reduced underage drinking driver fatality ratios. Trained servers are more likely to refuse service to visibly intoxicated patrons, and studies have documented decreases in patron blood alcohol levels and fewer calls to emergency services in establishments with trained staff.
The research does come with a caveat: training alone isn’t always enough. Reviews have concluded that server training is most effective when backed by strong management support and active enforcement. A trial in Sweden that paired training with enforcement reduced both over-service and violent assaults, while a similar program in Norway without the enforcement component did not. The takeaway for communities is that server training laws need teeth to deliver results.
Passive Alcohol Detection in New Cars
The most ambitious prevention technology on the horizon is the Driver Alcohol Detection System for Safety, a federal research program developing sensors that could be built into ordinary passenger vehicles. Unlike ignition interlocks, these wouldn’t require the driver to do anything. One version uses passive breath detection, sampling the air near the driver to estimate blood alcohol concentration. Another reads alcohol levels through the skin of a driver’s fingertip when they touch the start button or steering wheel.
As of the most recent federal progress report, breath sensors intended for widespread passenger vehicle use were targeted for production readiness around 2024, while touch-based sensors for privately operated vehicles were targeted for 2024 to 2025. These timelines represent hardware readiness, not mandatory installation in new cars, which would require additional regulatory and legislative steps. But the technology is no longer theoretical. Fleet vehicles and aftermarket accessories are expected to be the first applications, with broader adoption following as the sensors prove reliable in real-world conditions.
If passive detection eventually becomes standard equipment, it would fundamentally change the math on impaired driving by catching every impaired driver, not just the ones who encounter a checkpoint or get pulled over. Combined with the personal, community, and policy strategies already proven to work, it represents a future where 13,000 annual deaths is no longer treated as an acceptable cost of driving.