Drink-Driving Limits Run From Zero to 0.08% Around the World

In Australia a police officer can pull over any driver, on any public road, at any time, and ask for a breath sample. Nobody needs to be weaving, speeding or suspected of anything. That single rule explains why the Australian approach to drink driving differs so sharply from the American one, and why the same reading is policed so differently in each country.

What is the drink-driving limit in Australia?

The maximum blood alcohol content is 0.05% in every Australian state. A driver whose preliminary roadside reading is 0.05% or more is usually arrested and taken to a police station or a testing bus, where a more sophisticated machine produces a breath analysis. That result, unlike the roadside figure, can be used as evidence in court to prove a drink-driving charge.

The limit is enforced through random breath testing, which began in Victoria in July 1976. New South Wales launched its own big RBT campaign in 1982, and the share of deaths on the road that involve alcohol has since fallen: from roughly 40 per cent of all fatalities to 15 per cent by 2012. Police in New South Wales conduct about 5 million breath tests every year, and every police car in the state can carry out a mobile test.

In Australia, a police officer can stop any driver on any public road for a random breath test.

How do drink-driving limits compare around the world?

They run from zero to 0.08%, with a great many countries sitting at 0.05%. Zero tolerance is a real policy in places such as Bahrain, where drinking is legal but the limit is zero, so any alcohol above the limit of detection brings penalties. In the European list, "zero" generally means below the detection limit.

Within a country the limit often depends on who is driving. Canada's provinces can apply zero tolerance to new drivers under graduated licensing, or to those under 22. Austria drops to 0.01% for drivers licensed for less than two years and for vehicles over 7.5 tonnes, Croatia sets zero for drivers aged 18 to 24 and professionals on duty, and Thailand applies zero to professional drivers and 0.02% to new ones. Denmark's limit for the first three years after getting a licence is 0.02%, a rule justified by the thinner experience of newly licensed drivers.

Countries with no limit, and with total bans

World Health Organization data from 2018 listed eighteen countries with no legal limit at all, from Barbados and Burundi to Gambia, Guatemala, Liberia, Niger, Senegal, South Sudan and Togo. At the other end, Afghanistan, the Maldives, Mauritania, Saudi Arabia, Somalia, Sudan and Yemen had total bans covering every type of driver. Iran, Kuwait and Saudi Arabia ban alcohol outright.

How is blood alcohol measured, and what do the units mean?

Blood alcohol content measures how much alcohol, by mass, sits in a given volume of blood. America and many other countries quote it as a percentage, so 0.08% equals 0.8 g/L. The percentage means grams per deciliter, and the same level of 0.05% appears as 0.5 per mille or 50 mg% elsewhere. Reporting conventions vary by country, which is why one traveller's 0.05 can look like another's 0.5.

Police rarely draw blood at the roadside. A breathalyzer measures breath alcohol content, and jurisdictions apply a statutory conversion factor from breath to blood anywhere between 2000 and 2400. Japan, for example, sets its limit as 0.15 mg/L of breath, equivalent to 0.03%. Breath testing is the most common method in most jurisdictions, while blood tests rely on a venous sample analysed by forensic laboratories or hospitals. Germany measures the serum level and converts it to whole blood by dividing by 1.236, a calculation that underestimates BAC by 4% to 10% compared with other methods.

The law uses blood alcohol because it can be measured objectively and is difficult to contest in court, even though impairment can vary between people with the same reading. Most countries forbid operating motor vehicles and heavy machinery above prescribed levels, boats and aircraft are regulated too, and some jurisdictions also regulate cycling under the influence.

How do random breath tests differ from US sobriety checkpoints?

An Australian random breath test can happen anywhere on any public road, with no warning. A US sobriety checkpoint is a fixed operation: officers stop every vehicle, or every nth vehicle, at a set place, and the location is often announced in advance.

Australian operations involving a large number of officers, typically 10 to 20 at a fixed spot, are nicknamed booze buses. Police also fit mobile testing into ordinary patrols. Since the concept was extended in every state and territory, the same buses can double as random drug test buses that screen for cannabis, methamphetamine and ecstasy.

The American checkpoint

US checkpoints tend to be set late at night or in the early morning, on weekends and on party holidays such as New Year's Eve, when the proportion of impaired drivers is highest, and near the exits of public events. With a portable breathalyzer, police can test every driver where the law permits, processing cars one by one. Without one, an officer who notices the smell of alcohol or slurred speech asks the driver out for roadside field sobriety tests. These are not pass-or-fail: they only help the officer decide whether to arrest, after which the driver is asked for a breath or blood test. In the United States the field tests are voluntary.

The legal argument has been fierce. The Fourth Amendment guards against unreasonable searches and seizures, and a checkpoint stops drivers who are under no suspicion. The Michigan Supreme Court found sobriety roadblocks unconstitutional, but in Michigan Dept. of State Police v. Sitz in 1990 the US Supreme Court ruled 6-3 that properly conducted checkpoints are constitutional. Justice Stevens, dissenting, wrote that the net effect on traffic safety was infinitesimal and possibly negative. Some states, Washington among them, have held checkpoints unconstitutional under state law.

The police often announce a checkpoint in advance, by posting signs or notices in newspapers, and in the 2010s smartphone apps appeared that map reported checkpoints and warn drivers nearby. The stops catch more than drinkers: drivers hand over a licence, so officers run names for warrants, and can find unlicensed drivers, registration violations, drugs or weapons.

US law also treats drink driving as a special crime. Driving with a BAC over the set limit is the offence, so no one needs to drive recklessly or cause a crash to be convicted. Because proving BAC means a self-incriminating test, some jurisdictions set the penalty for refusing equal to, or worse than, the penalty for failing.

Canada and elsewhere

Canada runs the same idea under different names. In Ontario it is Reduce Impaired Driving Everywhere, or RIDE, in Alberta and Manitoba it is Check Stop, and in British Columbia it is Drinking Driving Counterattack.

What penalties follow a high reading?

Penalties differ by country. The menu includes fines, jail, a suspended licence, an impounded or seized car, and compulsory courses. The size of the reading usually decides which of these apply, and the same pattern of bands repeats from country to country.

In Mainland China a reading between 0.02% and 0.08% brings a fine of 1,000 to 2,000 yuan and a six-month suspension, while above 0.08% means up to three years' imprisonment and a five-year suspension. South Korea steps up from a 90-day suspension for 0.03% to 0.08%, to a one-year revocation for 0.08% to 0.20%, and finally to revocation for an indefinite period, up to 5 years in prison and a fine of up to KRW 20,000,000 above 0.2% or on a third catch. A drink-driving crash that kills carries a minimum of 3 years and up to life. Public anger at lenient sentencing led to harsher penalties there.

Chile separates driving under the influence (0.03% to 0.08%, a three-month suspension and a fine) from drunkenness (over 0.08%, a prison term of 61 to 301 days, a two-year suspension on a first offence, five years on a second and life on a third). Colombia, described as having the toughest penalties in Latin America, suspends the licence for a year for 20 to 39 mg/100 mL, and for the worst grade of 150 mg/100 mL, takes it for ten years with a fine of US$7,314. In India a first offence can bring six months' imprisonment, a 10,000 rupee fine or both, and a second within three years two years, a 15,000 rupee fine or both.

Denmark shows how steeply penalties can climb. For a first offence, 0.05% to 0.12% brings a fine and a conditional suspension, and the driver must retake the test and attend an alcohol-and-vehicles course. Between 0.12% and 0.20% the suspension is a full three years. Above 0.20% there is a conditional prison term of 20 days, a fine of a month's take-home pay, usually confiscation of the car, a three-year suspension and two years with a breath alcohol ignition interlock. Repeat offenders face significantly higher penalties at any level.

How is blood alcohol estimated from the number of drinks?

With a model from the 1920s by the Swedish professor Erik Widmark. It treats the body as a single compartment, assumes instant absorption and removes alcohol at a steady rate of about 0.15 g/L an hour. The volume the alcohol spreads through is body weight times 0.71 L/kg for men and 0.58 L/kg for women. The World Health Organization defines a standard drink as 10 grams of pure alcohol; the US standard is 14 grams.

The worked examples in the model show the scale. An 80 kg man who drinks 20 grams of ethanol is estimated at 0.204 g/L (0.0204%) after one hour, and a 70 kg woman who drinks 10 grams at 0.090 g/L (0.0090%). The model comes with caveats: it is most accurate a few hours after a single dose in a fasted state, can be within 20% of the true value, is not accurate below 0.2 g/L, and overestimates the peak when there is food in the stomach. Elimination commonly runs at 0.015 to 0.020 g/dL an hour, but metabolism varies with body temperature, the type of drink and the food eaten. Studies of trained volunteers found they could learn to judge their own BAC, with a typical error of 9 mg/100 mL (0.009%), but only after repeated breathalyzer feedback.

Courts also work backwards. Retrograde extrapolation projects a later test result back to the time of driving, and a growing number of states presume in law that the reading at the time of driving matched the later test, usually within two or three hours, though a defendant may offer evidence to rebut it.

Where is this heading?

Several recent changes tighten the rules. The Cayman Islands cut its limit to 0.07% from 0.10% in 2022. Argentina adopted zero tolerance in 2023: the law applies on national roads, and 18 of 24 provincial jurisdictions have adopted 0.0 g/l for provincial routes, while as of February 2024 six, including Mendoza and Santa Fe, kept their old limits. Malaysia's Road Traffic (Amendment) Act took effect on 23 October 2020, setting 22 micrograms per 100 millilitres of breath and 50 milligrams per 100 millilitres of blood.

Roadside enforcement is widening in scope too. The NSW Police Force announced in late 2017 that it would extend mobile drug testing to cocaine in 2018, with trials in Sydney's eastern suburbs.