A reasonable question, after seeing four different BAC limits cited on the same page (0.08% for most drivers, 0.05% in Utah, 0.04% for commercial drivers, 0.02% for drivers under 21), is how any of those numbers got picked. None of them are obvious. Blood alcohol content isn't a unit anyone has intuition for in the first place; the difference between 0.05% and 0.08% is about half a beer for an average adult. Picking which side of that gap counts as "impaired" is a policy decision dressed up as a scientific one, and it has a documented history.
The current US framework arrived in pieces. The federal government technically cannot set a per se driving limit at all, since DUI law is a state prerogative under the Tenth Amendment, but Congress can withhold highway funding from states that don't meet a target. Every BAC number on a state's books was set by that state's legislature. Most of them were set in response to a federal carrot or stick that came along at a specific date. The result is a layered system that looks coordinated and was actually negotiated state by state over roughly ninety years.
Where 0.15% came from, and why it didn't stay there
The earliest per se BAC threshold in US discussion was 0.15%. The figure traced to a 1938 American Medical Association committee and a 1939 joint recommendation issued with the National Safety Council, identifying 0.15% as the level above which most drinkers showed unambiguous laboratory deficits in coordination and judgment. "Per se" means presumed guilty at that level. The prosecutor doesn't have to demonstrate impairment beyond the test result. Before per se laws, drunk driving was prosecuted purely as impaired driving, with the BAC test serving as evidence rather than a definitive trigger.
The 0.15% number reflected mid-twentieth-century laboratory data on observable impairment, not lower-end safety risk. At 0.15%, slurred speech, reduced reaction time, and visible coordination deficits are reliably present in most drinkers. Below it, the picture gets noisier. The committee picked a number where the science was unambiguous, which was a defensible scientific position but not a public-safety one.
Through the 1960s and 1970s, the standard ratcheted down. The National Highway Traffic Safety Administration, established in 1970, used crash data rather than laboratory observation to ask the question differently: at what BAC level does crash risk start rising sharply against sober baseline? The answer was lower than 0.15%. Successive recommendations dropped the threshold to 0.10%, and most states had adopted that level by the early 1980s.
The 1980s push to 0.08%
Mothers Against Drunk Driving was founded in September 1980. Within five years the policy conversation had shifted from impairment thresholds to crash-fatality thresholds. The relative-risk curve for fatal crashes against BAC turns out to be approximately exponential. A driver at 0.08% is roughly four times as likely to be in a fatal crash as a sober driver; at 0.10%, about seven times; at 0.15%, more than twenty times. Setting the per se limit at 0.10% caught most fatal-crash drivers but missed a meaningful fraction in the 0.08% to 0.10% band.
Utah adopted 0.08% in 1983, becoming the first US state to fall below 0.10%. Other states followed gradually over the next decade, usually in response to MADD-led legislative campaigns. By the late 1990s, about nineteen states had adopted 0.08%, leaving roughly thirty-one still at 0.10%.
The federal lever changed the math. The Department of Transportation and Related Agencies Appropriations Act of 2001, signed by President Clinton on October 23, 2000, included a provision codified at 23 USC § 163 that directed the Secretary of Transportation to withhold a percentage of federal-aid highway funding from any state that did not enact a 0.08% per se BAC law by October 1, 2003. The withhold ramped: 2% of certain highway funds in fiscal year 2004, 4% in fiscal year 2005, with full restoration available retroactively if the state subsequently came into compliance. By 2004, every state had a 0.08% law on the books.
The same statutory mechanism, withhold a percentage of highway funds unless states comply, had been used five years earlier in the National Highway System Designation Act of 1995, which required states to enact zero-tolerance laws for drivers under 21 by October 1, 1998. The under-21 limits on state books today exist because of that statute. They aren't the product of fifty independent state policy decisions about youth drinking.
Why commercial drivers are held to 0.04%
The 0.04% federal limit for commercial drivers is older than the 0.08% limit for everyone else. It comes from the Federal Motor Carrier Safety Regulations at 49 CFR § 392.5, which prohibits anyone operating a commercial motor vehicle from doing so with a BAC of 0.04% or higher. The regulation was promulgated in the early 1990s under the authority of the Commercial Motor Vehicle Safety Act of 1986.
The 0.04% number wasn't picked by extrapolating the 0.08% standard. It was set independently, against a different reference: the level at which the FMCSA judged that the operation of a heavy commercial vehicle, with significantly more crash energy and longer stopping distances, became unacceptable from a public-safety standpoint. The threshold is half the per se driving limit, but the policy logic was "commercial operation tolerates less impairment," not "let's pick half of 0.08%."
The same 0.04% appears in the federal rules for airline pilots (14 CFR § 91.17) and merchant mariners (33 CFR § 95.020), though most carriers impose stricter operational rules. Pilots are typically prohibited from any alcohol within 8 to 12 hours of duty, regardless of measured BAC.
Utah's 0.05% break and what came next
Utah passed HB 155 in March 2017, and the law took effect on December 30, 2018, lowering the per se limit to 0.05% for non-commercial drivers. It was the first US state below 0.08% since the 2004 national compliance date. The policy rationale cited two anchors: the World Health Organization's 2010 recommendation of 0.05% as a global standard, and the NTSB's 2013 safety recommendation H-13-005, which urged all states to adopt 0.05%.
The data after Utah's transition has been examined in published research. A 2022 evaluation funded by NHTSA, the Insurance Institute for Highway Safety, and the Utah Department of Transportation found that fatal-crash rates dropped by roughly 20% in Utah in the year following the law's effective date, against a national fatal-crash rate that fell by about 6% over the same period. Critics argued that other factors (increased enforcement, public-awareness campaigns, broader trends in alcohol consumption) confounded the comparison. Proponents pointed to the same data as confirming what the relative-risk curve already implied: that meaningful fatal-crash reductions live in the 0.05% to 0.08% band, not in catching drivers above 0.15%.
The federal government has not used the highway-funding mechanism to push other states toward 0.05%. The NTSB recommendation remains an aspiration rather than a mandate. As of early 2026, no other state has moved below 0.08%, though Washington, New York, Hawaii, and Connecticut have all seen 0.05% bills introduced and stall in committee.
Where the US sits internationally
The 0.05% threshold Utah adopted is the international norm, not the outlier. The standard in most of Europe (France, Germany, Italy, Spain, Belgium, Netherlands, Austria, Switzerland, Portugal, Ireland) is 0.05%. Sweden uses 0.02%. Norway uses 0.02%. Japan uses 0.03%. The UK is split: England, Wales, and Northern Ireland sit at 0.08%, the only jurisdictions outside North America with a number that high; Scotland moved to 0.05% on December 5, 2014.
| Jurisdiction | Driver per se limit | Notes |
|---|---|---|
| Most US states | 0.08% | 50-state compliance by 2004 (23 USC § 163) |
| Utah | 0.05% | Effective December 30, 2018 (HB 155) |
| Scotland | 0.05% | Effective December 5, 2014 |
| England, Wales, Northern Ireland | 0.08% | Unchanged since 1967 |
| Most of EU | 0.05% | Phased in through 1990s–2000s |
| Sweden, Norway, Poland | 0.02% | Sweden since 1990, Norway since 2001 |
| Japan | 0.03% | Lowered in 2002 |
| Saudi Arabia, UAE | 0.00% (zero tolerance) | Religious and statutory basis |
The chart public-health agencies cite to justify pushing the US to 0.05% puts the 0.08% bracket in red and notes that the United States is one of three remaining OECD countries at the higher number. The chart opponents cite shows US crash-fatality rates have fallen by more than half since 1982 under the existing 0.08% framework, and argues that further reductions might come more cheaply from enforcement than from a threshold change.
Both charts are honest. They're answering different questions.
What the NTSB has been asking for since 2013
The National Transportation Safety Board first recommended a national 0.05% per se BAC limit in May 2013 (Safety Recommendation H-13-005), and renewed the recommendation on its Most Wanted List in 2022. NTSB recommendations are advisory only. The Board has no rulemaking authority. But its recommendations historically have a moderate adoption rate over the following decade, and the 0.05% recommendation has stayed on the list every year since it was issued.
The political path forward would be the same as the 2000 path: attach a per se BAC condition to a future federal highway-funding reauthorization, and let the states comply within a few years. There's no indication in current legislative drafts that this is being prepared, but there's also no precedent for the federal level standing still on this question for as long as it has. The roughly 14-year gap between NHTSA's first push for 0.08% as a federal target in the mid-1980s and the 2000 statute that made it the de facto national requirement is one reference point for how slowly the policy moves when it isn't actively being driven.
The day-to-day relevance for someone making a decision about whether to drive: the legal threshold is 0.08% in 49 states and 0.05% in Utah, and the gap between those two numbers is real. For an average-weight adult, 0.05% works out to roughly one fewer drink than 0.08% over the same window. The calculator math doesn't change with geography. The policy threshold sits a little tighter in Utah, and almost certainly will in more states before the decade is out.
Sober shows the legal threshold next to your estimated BAC and the time until you're below it, so the per se limit shows up where it actually matters: between the result and the decision.