The impairment behind millions of preventable crashes
Alcohol has been recognised as a road safety threat for as long as cars have existed — the world’s first drink-driving conviction predates the twentieth century. More than a hundred years later, driving under the influence of alcohol and, increasingly, drugs remains one of the leading contributors to road deaths worldwide. It is also among the most solvable: few areas of road safety have a clearer evidence base for what works.
This page explains how impairment kills, where the world sets its limits, and which countermeasures actually reduce deaths.
What counts as impaired driving?
- Drink-driving: operating a vehicle with a blood alcohol concentration (BAC) above the legal limit — or, in zero-tolerance jurisdictions, with any alcohol at all.
- Drug-driving: driving while impaired by illicit drugs (such as cannabis, cocaine, amphetamines or opioids) or by psychoactive medicines. Some countries prohibit any detectable amount of certain substances; others require proof of impairment.
- Combined use: alcohol taken together with drugs — including some prescription medicines — multiplies impairment beyond either substance alone, and features disproportionately in fatal crashes.
Why impairment kills: the science
Alcohol degrades almost every capability that driving depends on: reaction time, hazard perception, divided attention, judgement of speed and distance, and — critically — the drinker’s own ability to recognise how impaired they are.
Two findings from decades of research matter most:
1. There is no safe threshold — risk rises from the first drink. Measurable impairment of driving-related skills begins at BAC levels well below most legal limits. Crash risk then climbs steeply and non-linearly: relative to a sober driver, the risk of a fatal crash is elevated at 0.05 g/dL, several times higher at 0.08 g/dL, and rises dramatically beyond that. This exponential curve is why even “moderate” drink-driving is far more dangerous than intuition suggests.
2. Young and novice drivers are hit hardest. Inexperience and alcohol interact viciously: at the same BAC, a young novice driver faces a far higher crash risk than an experienced adult. This is the evidence behind the lower or zero limits that many countries apply to new drivers.
Drugs impair differently by substance — cannabis slows reaction time and distorts distance perception; stimulants encourage risk-taking and aggressive driving; opioids and sedative medicines cause drowsiness — but the overall finding is consistent: psychoactive substances degrade driving performance, and combining them with alcohol is especially lethal.
The scale of the problem
The World Health Organization identifies drink-driving as a key risk factor in road traffic deaths globally. The share of deaths involving alcohol varies widely by country — commonly cited figures range from under 10% in some jurisdictions to a quarter or more of all road deaths in others — with the true share often higher than official statistics show, because testing of drivers after crashes is far from universal.
Drug-driving is a growing concern almost everywhere, driven by changing patterns of cannabis use and legalisation in some jurisdictions, the spread of stimulants, and rising use of psychoactive medicines in ageing populations. Reliable data lag well behind alcohol: many countries only test crash-involved drivers for drugs inconsistently, if at all.
Where the world sets its limits
Legal BAC limits for general drivers cluster into a few groups:
| Approach | Typical limit | Examples of jurisdictions using this approach |
|---|---|---|
| Zero / near-zero tolerance | 0.00–0.02 g/dL | Several Eastern European, Middle Eastern and Asian countries |
| Low limit | 0.03–0.05 g/dL | Most of continental Europe (0.05), Japan (0.03), Sweden and Norway (0.02) |
| Higher limit | 0.08 g/dL | The United States (most states), Canada, England and Wales |
Two features of good practice recur internationally: lower or zero limits for novice and professional drivers (adopted across most of Europe and beyond), and limits set at 0.05 g/dL or below for the general driving population — the level long recommended by international road safety bodies, based on the risk curve described above.
What works: the evidence-based toolkit
1. Random breath testing (RBT)
The single most powerful tool against drink-driving is highly visible, unpredictable, high-volume enforcement — stopping drivers without needing suspicion. Australia’s decades-long RBT programmes are the classic evidence: sustained, mass testing produced large and lasting reductions in alcohol-related deaths. The mechanism is general deterrence — changing the perceived likelihood of being caught for everyone, not just catching offenders. Jurisdictions restricted to “suspicion-based” testing consistently achieve weaker results.
2. Lower BAC limits — properly enforced
Reducing the general limit (for example from 0.08 to 0.05) reduces alcohol-related deaths, with the effect strongest where the change is paired with visible enforcement and publicity. Zero-tolerance limits for young and novice drivers have repeatedly been shown to cut fatal crashes in that group.
3. Alcohol interlocks
Breathalyser-immobilisers fitted to offenders’ vehicles reliably prevent reoffending while installed, and are increasingly used in rehabilitation programmes across Europe, North America and Australasia — particularly for repeat offenders, for whom licence bans alone often fail (a substantial share of banned drivers simply keep driving). The EU now requires new car types to be pre-equipped with a standardised interlock installation interface. The frontier technology is passive, universal alcohol detection built into all vehicles — under active development and the subject of regulatory mandates being phased in the United States.
4. Swift, certain, graduated penalties
As with speeding, evidence favours certainty over severity: a high perceived chance of detection deters more than draconian but rarely applied punishments. Effective systems combine immediate licence suspension, graduated sanctions by BAC level and repeat status, mandatory assessment or rehabilitation for high-BAC and repeat offenders, and interlock-conditional relicensing. Treating the highest-risk group — drivers with alcohol dependence — as a health problem as well as an enforcement problem measurably reduces reoffending.
5. Drug-driving detection
Roadside oral fluid (saliva) screening for common drugs, now deployed in Australia, much of Europe and elsewhere, extends the deterrence logic of RBT to drugs. The policy design questions are harder than for alcohol — per se limits versus impairment standards, how to treat medicinal use, and detection windows that outlast impairment for some substances (notably cannabis) — but the direction of travel worldwide is firmly toward routine roadside drug screening.
6. Reducing exposure
Measures that separate drinking from driving in the first place: late-night public transport, ride-hailing availability, server liability and responsible-service rules, and designated-driver norms. These work best as complements to enforcement, not substitutes.
Common myths, briefly answered
“Coffee, food or a cold shower sober you up.” Nothing but time reduces BAC; the body eliminates alcohol at a roughly fixed rate. Feeling more alert is not being less impaired.
“I drive more carefully after a few drinks, so it evens out.” One of alcohol’s first casualties is self-assessment: impaired drivers systematically underrate their impairment. The compensation drivers believe they are applying is itself degraded.
“The morning after is fine.” After heavy drinking, BAC can remain above legal limits well into the next day. A significant share of drink-driving detections occur in the morning.
“Cannabis makes you a slower, safer driver.” Cannabis measurably impairs reaction time, tracking and attention. The observed effect on crash risk is smaller than alcohol’s at typical levels — but it is an increase, not a decrease, and combining cannabis with alcohol sharply multiplies risk.
“Drink-driving laws punish social drinkers, not the real problem drivers.” Both matter. High-BAC and repeat offenders are disproportionately represented in fatal crashes — which is why interlocks and rehabilitation target them — but the far larger population of moderate drink-drivers also contributes substantially to total deaths, which is what general deterrence addresses.
The bottom line
Impaired driving is the risk factor where the world knows most precisely what to do: set limits at 0.05 g/dL or below with zero tolerance for novices and professionals, enforce them through mass random testing that makes detection feel likely, fit interlocks to offenders’ vehicles, treat dependence as the health condition it is — and extend the same deterrence architecture to drugs. Every element of that package has been evaluated in the real world and shown to save lives. The gap between the best-performing countries and the rest is not a gap in knowledge. It is a gap in implementation.
Sources and further reading: World Health Organization — Global Status Report on Road Safety and drink-driving road safety manual; International Transport Forum (ITF/OECD) publications on alcohol, drugs and driving; European Transport Safety Council (ETSC) — “Safe & Sober” programme publications; national evaluations of random breath testing (Australia) and alcohol interlock programmes (Sweden, Finland, EU). Figures cited are drawn from this body of international research; see each organisation’s latest publications for current data.
Part of the Road Safety News Risk Factors series: [Speeding] · Drink & drug driving · Seat belts & child restraints · Helmets · Distracted driving · Fatigue.
