Alcohol testing accurately predicts driver impairment. But drug testing works differently, and can be positive long after the effect of the drugs have worn off.
I have seen drug use up close
My academic work examines how law, evidence and politics intersect in the regulation of drugs. But my understanding of this subject is not drawn solely from books and journal articles. Before entering academia, I worked as a nightclub bouncer, and long before that, as a young person navigating environments where drug use was neither abstract nor theoretical. I have seen drug use from the dance floor, from the street, and from the courtroom. That lived experience matters. It means I have witnessed the difference between impairment and presence. I have seen individuals clearly under the influence and others who were not. I have seen chaotic use, functional use, recovery, relapse, and everything in between. I have also seen how the law often struggles to reflect that complexity.
Drug policy in the UK is frequently framed as evidence-based and harm-focused. Yet when we examine how certain offences are constructed, particularly drug driving, uncomfortable questions emerge about whether the science genuinely supports the legal thresholds being applied.
This is not an argument for leniency. It is an argument for accuracy, proportionality, and justice.
Punished for yesterday
We are repeatedly told that drug driving laws exist to keep us safe. To protect the public. To remove dangerous drivers from our roads. It is a reassuring narrative firm, scientific, evidence-based.
But what if the science is not nearly as firm as we have been led to believe?
Recently, a man appeared in court charged with drug driving. He had been stopped at the roadside for a random road safety check, he was polite and cooperative. There were no signs of impairment. He was not swerving, he was not incoherent, there was no erratic behaviour. When asked if he would mind doing a drink and drugs test he said of course, as he didn’t think there would be a problem. What he did have in his blood, which he didn’t know, was benzoylecgonine, the primary metabolite of cocaine.
He was in recovery from cocaine use, he had not used for several days and so assumed he was safe to drive.
Benzoylecgonine is not cocaine, it is an inactive breakdown product. It does not produce euphoria, does not stimulate, does not impair, it is simply what remains after the body has processed the drug. Yet under Section 5A of the Road Traffic Act 1988, introduced via the Crime and Courts Act 2013, its presence above a specified threshold is enough to constitute a criminal offence.
He was not being prosecuted for driving while impaired.
He was being prosecuted for having used drugs in the past.
The difference between alcohol and drugs
With alcohol, the law is built on decades of research demonstrating a relatively consistent relationship between blood alcohol concentration (BAC) and impairment. Reaction time slows. Judgement deteriorates. Crash risk increases in a dose-related way. Legal limits were introduced because crash-correlation data showed increased danger at particular concentrations.
The same clarity does not exist for most controlled drugs.
Cocaine is a short-acting stimulant. Its psychoactive effects typically last one to three hours. After that, it is metabolised into benzoylecgonine, which can remain detectable in blood for several days longer, in chronic users up to 14. Benzoylecgonine has no psychoactive effect, yet UK law sets a specified limit for it.
What robust evidence demonstrates that exceeding that limit equates to impaired driving ability at the time of driving? The evidence base is far thinner than the public is often led to believe.
Drugs cause less traffic problems than alcohol
Professor David Nutt, former Chair of the Advisory Council on the Misuse of Drugs (ACMD), has repeatedly highlighted the weak relationship between blood concentrations of many drugs and actual behavioural impairment. In his widely cited study in The Lancet, Nutt, King and Phillips (2010) found that alcohol causes greater overall harm than many illegal substances, including heroin and crack cocaine. Yet alcohol remains legally regulated through calibrated impairment thresholds, while illicit drugs are treated as inherently criminal regardless of comparative harm.
The issue is not whether drugs can impair, some clearly can. The problem is that, unlike alcohol, there is often no simple, linear relationship between blood concentration and functional impairment. Tolerance varies, frequency of use matters and chronic users metabolise substances differently. Detection windows frequently exceed impairment windows.
Similarly to the case of benzoylecgonine with cocaine use, in the case of cannabis, Nutt and other researchers have argued that blood THC levels correlate poorly with impairment, especially in regular users. Impairment may occur at relatively low concentrations shortly after use, yet measurable traces can persist long after functional ability has returned to baseline. A 2020 randomised clinical trial published in JAMA Psychiatry found that while THC can impair driving performance, blood levels alone were not reliable indicators of functional impairment.
The pharmacology does not map neatly onto the legal threshold.
The limits were not built like alcohol limits
When drug driving limits were introduced in 2014, the Government presented them as “evidence-based”. However, the ACMD’s own 2013 report on driving under the influence of drugs acknowledged that limits were derived from a combination of risk modelling, international comparisons, and pragmatic enforcement considerations – not decades of crash-correlated impairment validation equivalent to alcohol research.
This distinction matters.
Criminal law carries stigma, mandatory disqualification, financial penalties, and in some cases imprisonment. To justify that level of sanction, we should expect robust evidence that the prohibited concentration reliably indicates dangerous incapacity.
Instead, in cases involving inactive metabolites such as benzoylecgonine, what is being measured is not impairment but biological trace.
That is a profound shift in the logic of criminal responsibility.
A broader scientific problem
The difficulty runs deeper than drug driving alone.
The classification of drugs in the UK has long been criticised for failing to align with relative harm. Nutt himself was removed as Chair of the ACMD in 2009 after publicly stating that alcohol and tobacco were more harmful than certain illegal drugs, including ecstasy. His later research reinforced the conclusion that alcohol is one of the most harmful drugs overall when considering both individual and societal damage.
If harm alone determined legality, alcohol would not sit comfortably outside the Misuse of Drugs framework.
Instead, all illicit drug use is criminalised within a system that has never been coherently structured around comparative scientific harm. That framing drugs as inherently problematic regardless of context, dose or timing now shapes drug driving law.
Presence substitutes for impairment. history substitutes for risk.
Civil liberties and the presumption of innocence
If the science underpinning thresholds is contested, enforcement policy is now pushing even further.
Senior police representatives, including the National Police Chiefs’ Council roads policing lead, have publicly called for powers to allow officers to impose immediate roadside bans on suspected drink and drug drivers before court proceedings are concluded.
The Government has also consulted on proposals for temporary licence suspension powers for drivers arrested on suspicion of drink or drug driving while awaiting court.
On the surface, these measures are framed as pragmatic road-safety tools. But they represent a significant constitutional shift. In England and Wales, the presumption of innocence until proven guilty beyond reasonable doubt is a foundational principle of criminal justice. Disqualification has historically followed conviction, not suspicion. To impose punitive consequences before judicial determination blurs the line between administrative precaution and criminal punishment.
When combined with thresholds that do not necessarily reflect impairment, the risk is clear, suspicion plus residue becomes sanction. Public safety is essential, but it should not come at the expense of evidential integrity and due process.
What are we really criminalising?
In conclusion , none of this is an argument for allowing impaired drivers on the road. Impairment from alcohol, drugs, fatigue or distraction is dangerous and should be addressed robustly.
The question is whether the current framework accurately distinguishes danger from detection.
When inactive metabolites trigger conviction, the law is no longer measuring impairment – it is measuring chemistry. When thresholds lack clear crash-risk validation, scientific language risks becoming a veneer rather than a safeguard. And when proposals edge toward penalties before conviction, the presumption of innocence begins to look conditional.
This is not simply a technical flaw, it is a philosophical one.
Criminal law should punish harmful conduct. It should not punish category membership. It should respond to demonstrable risk, not historical consumption.
So we must ask, plainly, are we criminalising dangerous driving or are we criminalising people for what remains in their bloodstream long after danger has passed? Criminalising them for their social choice to use illicit drugs?
Friends of Bylines Network
There has never been a greater need for grassroots journalism that investigates the stories that really matter, holds power to account and champions the voices of everyday citizens. We are proudly powered by volunteers but what we do isn’t free.
STAND WITH US for independent, citizen-led journalism that makes democracy stronger, and you will even get some exclusive benefits.
