The Brands With America’s Drunkest Drivers Are the Expensive Ones

According to the toxicology reports, the mental image most Americans carry around of impaired driving is wrong. Not the severity. Not the body count. The vehicle. We aggregated FARS toxicology results across 490,736 drivers involved in fatal crashes from 2014 to 2023, grouped them by brand, and the leaderboard looks nothing like the stereotype^[1].

Infiniti. Nearly one in four Infiniti drivers in a fatal crash tested positive for alcohol or drugs. Behind them: Pontiac at 22.4%, Audi at 22.0%, Mercedes-Benz at 21.9%, Volvo at 21.5%^[1]. Four of the top six are luxury or near-luxury. Pontiac is the outlier, but its fatal-crash fleet skewed heavily toward the Grand Prix, Grand Am, and GTO, all marketed as performance vehicles to buyers who shared more in common with Infiniti G37 shoppers than Pontiac Vibe owners. Meanwhile, Toyota sits at 19.1%. Ford at 19.3%. Lincoln at 19.0%. America’s workhorses are, statistically, America’s soberest.

A separate 2024 analysis by LendingTree found the same pattern approaching from the other direction: BMW owners accumulated 3.09 DUI violations per 1,000 drivers, the highest of any brand. Acura clocked 2.7. Audi landed at 2.4^[2]. Two completely different datasets. Same conclusion.

Within the Same Brand, the Spread Is Enormous

Brand averages, though, obscure something wilder. The gap between the drunkest and soberest brands is 5.4 percentage points. The gap within a single brand can hit 21.

Toyota’s range: Solara at 4.1% impairment to FJ Cruiser at 25.3%. Cadillac: Seville at 10.5% to CTS at 25.9%. Subaru: Ascent at 8.2% to WRX at 23.4%^[1]. Pick any Toyota off the lot and your impairment-in-fatal-crash odds swing by a factor of six depending on which model you chose. The badge on the steering wheel is a weaker predictor of who’s driving sober than the specific nameplate underneath it.

What This Actually Measures

Methodology matters. FARS captures fatal crashes only. Toxicology testing rates vary by state, and “impaired” here means BAC above zero or drug-positive, which includes trace amounts that may not indicate active impairment at the time of the crash^[1]. A driver who smoked marijuana three days earlier and drove sober counts the same as someone at 0.20 BAC.

So these numbers overstate impairment in one sense and understate it in another. What they do reliably capture is the relative ranking. If Infiniti drivers are testing positive 5.4 points above Toyota drivers under the same flawed methodology, the methodology isn’t the explanation.

The Strongest Case Against This Finding

Demographics. Luxury brand buyers in fatal crashes may skew younger than their brand image suggests. Performance-oriented models (G37, CTS-V, WRX) attract younger drivers, and younger drivers across every brand carry higher impairment rates. Strip out drivers under 30 and some of the luxury premium might evaporate. The brand could be acting as a proxy for age and risk appetite rather than income. A 2012 UC Berkeley study did find drivers of expensive cars were less likely to yield to pedestrians at crosswalks^[3], suggesting wealth and entitlement play some role. But the FARS data alone cannot untangle age from income from vehicle choice.

What the data can say with confidence: brand-level differences are real but modest (19% to 24%). Model-level differences within brands dwarf them (4% to 25%). Your neighbor’s Ascent and your neighbor’s WRX share a Subaru badge and almost nothing else in the toxicology column. If you want to predict who’s driving sober, stop looking at the logo and start looking at the trim level.