Cannabis vs Alcohol on Australian Roads: What the Data Actually Shows

Last updated: January 2026. This article is intended for educational purposes. If you hold a medicinal cannabis prescription, consult your prescribing clinician before driving.

Why This Comparison Matters

When a coronial finding or roadside drug test result appears in Australian media, the framing is often stark: cannabis is on our roads and people are dying. What those headlines rarely provide is denominator data — how many of those deaths involved cannabis as a contributing cause versus cannabis simply being present in a deceased person's toxicology. Alcohol, by contrast, has decades of dose-response research firmly establishing causal impairment thresholds.

This article compiles peer-reviewed research, Australian Institute of Health and Welfare (AIHW) data, National Coronial Information System (NCIS) figures, and international pharmacological evidence to give a proportionate picture of what cannabis and alcohol each contribute to road trauma in Australia. It does not minimise cannabis driving risk — any impairment while driving is unacceptable — but it does insist that evidence, not media volume, should shape policy and public understanding.

Alcohol and Road Deaths: The Established Burden

Alcohol's role in Australian road trauma is large, well-measured, and causally understood. Key data points include:

• The AIHW's Alcohol, tobacco & other drugs in Australia report (2024 edition) identifies alcohol as a factor in approximately 24–30% of all road crash fatalities nationally, consistent across the 2015–2022 period.
• The Bureau of Infrastructure and Transport Research Economics (BITRE) estimated in its 2022 road safety analysis that alcohol-related crashes cost Australia approximately $5.7 billion annually in social and economic costs.
• Blood alcohol concentration (BAC) above 0.05 g/100 mL doubles crash risk; above 0.15 g/100 mL, crash risk increases by a factor of roughly 25, based on international meta-analyses replicated in Australian conditions.
• Random breath testing (RBT) data from state police services consistently detects alcohol over the legal limit in a meaningful proportion of serious-crash drivers — New South Wales alone conducts more than three million RBT stops per year.

These figures are credible because they rest on a validated dose-response framework. We understand the biology of ethanol, we have a legal impairment threshold (0.05 BAC in Australia), and we can trace cause from exposure to outcome.

Cannabis in Toxicology Reports: Presence Is Not Impairment

This is the single most important methodological point in the cannabis–driving literature, and it is routinely omitted from news coverage.

THC (delta-9-tetrahydrocannabinol), the primary psychoactive compound in cannabis, is detectable in blood for hours after acute intoxication has resolved. Its primary metabolite, THC-COOH, is detectable in urine for days to weeks after last use in regular consumers — without any ongoing psychoactive effect. When a coroner records "cannabis detected in toxicology," that finding alone does not establish that cannabis contributed to the crash.

A landmark systematic review published in the British Medical Journal (Rogeberg & Elvik, 2016) and a subsequent Cochrane-adjacent meta-analysis found that studies failing to adjust for alcohol co-use and other confounders consistently overestimated cannabis's independent contribution to crash risk. When alcohol was statistically controlled, the odds ratio for cannabis-only impairment fell substantially — and in several studies became non-significant.

That is not a finding that cannabis is safe to drive on. It is a finding that the literature has a confounder problem, and that unadjusted toxicology presence figures should not be used as though they were impairment figures.

Australian-Specific Crash Data

The most comprehensive Australian dataset examining drug-driving comes from the NCIS and state-level coronial studies. Key findings from the published literature:

Across these datasets, cannabis presence rates in road fatalities have remained broadly stable at around 10–13%. Critically, the studies that examined culpability — whether the driver with the detected substance was actually at fault for the crash — consistently showed that cannabis-positive drivers were judged culpable at lower rates than alcohol-positive drivers, and that cannabis-only drivers (no alcohol co-detected) showed culpability rates closer to drug-free drivers than to alcohol-impaired drivers.

This does not mean cannabis is safe for driving. It means that aggregate "cannabis detected" figures in media reporting substantially overstate attributable risk compared with what culpability-adjusted analyses find.

What Does Cannabis Actually Do to Driving Performance?

Laboratory and driving-simulator studies offer a more controlled picture than crash statistics. The consistent findings are:

• Acute THC impairment is real. Doses equivalent to recreational use impair divided attention, processing speed, and lane-keeping for approximately 3–4 hours after inhalation, and longer after oral/edible consumption (4–8+ hours depending on dose).
• Impairment is dose- and route-dependent. Low doses in occasional users produce more measurable impairment than the same nominal dose in heavy chronic users, due to tolerance — though tolerance does not fully protect driving performance.
• Cannabis and alcohol combined are far worse than either alone. The interaction is supra-additive for most driving-relevant cognitive measures. The strongest harm-reduction message around cannabis and driving is also an alcohol message: never combine them before driving.
• Oral/edible cannabis (including many medicinal preparations) has delayed onset and prolonged duration. Users may feel little effect for 1–2 hours and then experience significant impairment well after they believed themselves capable of driving.

Roadside Drug Testing in Australia: What It Measures

All Australian states and territories conduct roadside drug testing (RDT) for THC using oral fluid (saliva) swabs. It is important to understand what these tests detect:

• Oral fluid RDT tests for THC itself, not metabolites — this means detection windows are shorter than urine (generally 4–12 hours for occasional users, up to 24 hours for heavy users).
• A positive RDT result confirms recent cannabis use, but Australian law does not require proof of impairment. Presence of THC is the offence under state road transport legislation in every jurisdiction.
• This zero-tolerance legal model differs from alcohol, where impairment is inferred from a calibrated threshold. The policy rationale is precautionary, not pharmacological.
• Medicinal cannabis patients are not exempt from RDT laws in any Australian state or territory as of January 2026. A valid prescription does not provide a legal defence to a positive oral fluid test result. Patients should discuss this explicitly with their prescribing clinician.

Contextualising Media Coverage: A Methodological Checklist

When you encounter a news article or social media post claiming cannabis is "causing" a certain percentage of road deaths, apply these questions:

• Is this presence or attribution? "Detected in toxicology" is not the same as "caused or contributed to the crash."
• Was alcohol controlled for? Co-detection rates of 4–6% mean a significant share of "cannabis-positive" crash deaths also involved alcohol, which may be the operative impairment.
• Is there a culpability analysis? Studies that examine fault-assignment produce very different conclusions from raw detection counts.
• What is the denominator? Absolute counts of cannabis-detected fatalities mean little without the total crash population and cannabis use prevalence in the driving population.
• What time period does the data cover? Medicinal cannabis prescriptions in Australia grew from fewer than 3,000 active approvals in 2018 to over 400,000 by 2024 (TGA data). Any trend analysis must account for changes in population prevalence.

Harm Reduction: Practical Guidance

The evidence base does not support the conclusion that cannabis is harmless for drivers. The appropriate response to the nuanced data above is informed harm reduction, not dismissal of risk.

• Wait at least 4–6 hours after inhalation before considering driving, and longer if you feel any residual effect. Many clinicians and the TGA product information for medicinal cannabis products recommend longer waiting periods.
• For oral/edible/oil preparations, wait at least 8–12 hours, or as directed by your prescribing clinician — peak impairment may occur well after consumption and persist longer than expected.
• Never drive if you feel impaired. Tolerance does not reliably prevent all driving-relevant impairment, and self-assessment of impairment is unreliable under cannabis.
• Never combine cannabis with alcohol before driving. This is the highest-risk combination in the literature.
• Medicinal cannabis patients should consult their prescribing doctor about specific waiting periods for their formulation, dose, and individual circumstances. State-based guidelines vary and are subject to change.

Summary

Alcohol is responsible for a larger, better-characterised, and causally stronger share of Australian road fatalities than cannabis — consistently around 25–30% of crash deaths versus 10–13% cannabis detection rates that do not equate to attribution. The cannabis figure is further complicated by a detection methodology that does not map cleanly to impairment, widespread alcohol co-detection, and the absence of a validated impairment threshold equivalent to BAC.

None of this means it is safe to drive after using cannabis. It means the public conversation should be grounded in methodology, not media volume — and that harm reduction messaging should be proportionate, evidence-based, and directed at the behaviours that carry the highest actual risk.

Sources

• Australian Institute of Health and Welfare. Alcohol, tobacco & other drugs in Australia (2024). https://www.aihw.gov.au/reports/alcohol/alcohol-tobacco-other-drugs-australia
• Bureau of Infrastructure and Transport Research Economics (BITRE). Road safety: road deaths data portal. https://www.bitre.gov.au/statistics/safety/fatal_road_crashes_database
• Therapeutic Goods Administration. Access to medicinal cannabis: authorised prescribers and SAS statistics. https://www.tga.gov.au/products/unapproved-therapeutic-goods/medicinal-cannabis-hub/medicinal-cannabis-access-data
• Drummer OH et al. (2004). The involvement of drugs in drivers of motor vehicles killed in Australian road traffic crashes. Accident Analysis & Prevention, 36(2), 239–248. https://doi.org/10.1016/S0001-4575(02)00167-4
• Rogeberg O & Elvik R. (2016). The effects of cannabis intoxication on motor vehicle collision revisited and revised. Addiction, 111(8), 1348–1359. https://doi.org/10.1111/add.13347
• National Coronial Information System (NCIS), Monash University. https://www.ncis.org.au
• Transport for NSW. Drug driving. https://www.transport.nsw.gov.au/road-safety/driver-safety/drug-driving
• Arkell TR et al. (2021). Detection of impairment and THC in blood and oral fluid following acute cannabis use in medicinal cannabis patients. Psychopharmacology, 238, 3163–3173. https://doi.org/10.1007/s00213-021-05915-7