A tolerance that was never a legalisation
Nothing was ever legalised. Dutch policy from the mid-nineteen-seventies separated two categories of substance and declined to prosecute retail sale of one of them, under conditions, while leaving everything about it formally prohibited.
That arrangement produced a visible retail sector supplied by a chain that remained entirely illegal — a contradiction the Dutch themselves named, discussed for decades, and only began to address in the last few years.
This page covers what the policy actually said, what the conditions are, why the supply problem was structural, how the sector has shrunk, what the current experiment is doing, and what none of it has to do with the compliant European market.
⚠️ Scope. This page describes a process and what can be verified from documents. It does not describe what a product does to a person, gives no amounts, and does not replace a healthcare professional.
🔴 What the policy actually is
The legal position
The underlying legislation prohibits the substances. Nothing in the arrangement changed that, and possession and sale remained offences throughout.
The prosecutorial decision
Guidelines instructing prosecutors not to pursue certain offences under specified conditions. It is a decision about enforcement priorities rather than a change in the law.
The Dutch term
The policy has its own word, translating roughly as a policy of tolerance. The vocabulary itself makes clear that something prohibited is being permitted rather than made lawful.
Why the distinction matters
Because a tolerance can be withdrawn administratively, and a legalisation cannot. The sector has always operated at the discretion of the authorities.
The 1976 separation
What changed
The legislation was revised to distinguish substances presenting unacceptable risk from those considered to present lower risk, with different enforcement approaches for each.
The reasoning
To separate the markets. The argument was that a user of a lower-risk substance encountering a dealer who also sold higher-risk ones was the greater harm.
What it produced
A retail channel for one category, deliberately separated from the market for the other. That separation was the policy’s central objective.
How it was assessed
Dutch and international evaluations have generally found the market separation objective substantially achieved, whatever else is disputed.
The conditions
No advertising
Premises may not promote themselves. This is why the sector’s shopfronts are visually restrained by the standards of any other retail category.
Nothing from the other category
Selling substances from the higher-risk list ends the tolerance immediately. It is the condition the whole policy exists to enforce.
No nuisance
Premises causing disturbance to their surroundings lose their position. It gives local authorities substantial leverage.
No minors
A strict age threshold, enforced at the door. Breaches are treated as serious.
The quantity conditions
Per transaction
A small limit per customer per visit, set in the guidelines. It is intended to stop onward distribution.
On the premises
A limited stock at any one time, well below what a day’s trade requires. It is the condition that creates the daily supply problem described below.
Why the stock limit matters
Because premises must be resupplied frequently, from sources that are formally criminal, in transactions that are not tolerated.
The practical result
A legal fiction operating on a daily basis, which every participant and every authority understood perfectly well.
🔴 The supply contradiction
The Dutch name for it
The problem of the back door. Sale at the front is tolerated; supply at the back is not, and both are necessary for the premises to function.
What that meant in practice
An entire supply chain — cultivation, transport, wholesale — remaining fully criminal while supplying a tolerated retail sector.
The consequences
No quality control, no taxation of the supply chain, no product standards, and organised crime occupying the whole upstream.
How long it persisted
From the beginning. It was identified immediately, discussed continuously, and left unresolved for around four decades.
The residence condition
What was introduced
A criterion restricting access to residents, adopted in the early twenty-tens after pressure over cross-border traffic.
How it has been applied
Unevenly. Enforcement was left substantially to local authorities, and border municipalities have applied it far more strictly than others.
What it was for
Reducing the volume of visitors travelling specifically for access, which had become a significant issue in towns near the borders.
Its status
It remains part of the criteria, applied at local discretion. The variation between municipalities is considerable.
The sector has been shrinking
The peak
The number of tolerated premises peaked in the nineteen-nineties, at well over a thousand nationally.
Now
Substantially fewer, in the region of half that figure, and the decline has been steady rather than abrupt.
Why
Municipal policy, principally: distance requirements from schools, refusal to replace closed premises, and stricter application of the conditions.
What that reflects
A policy that was never expansionary. Local authorities have consistently used their discretion to reduce rather than to grow the sector.
Amsterdam specifically
Its share
A minority of the national total, concentrated in the centre, and much reduced from its peak.
The recent direction
Restrictions aimed at reducing tourism pressure on the centre, including measures affecting where and how the premises operate.
The city’s position
Repeatedly stated as wanting fewer visitors coming for this specific reason, which is close to the opposite of the city’s international reputation.
What that means
The place most associated with the policy internationally has been actively reducing its visibility for years.
The current experiment
What it is
A controlled trial establishing a legal supply chain: licensed producers supplying participating premises in selected municipalities, with the product tracked.
The legal basis
Legislation adopted in 2019 creating a framework for the trial, with a defined duration and an evaluation requirement.
What it addresses
The back door. For the first time, supply to participating premises is lawful and traceable rather than criminal.
Where it stands
A transition period ran from the mid-twenty-twenties, with participating municipalities moving to regulated supply. It is a trial with an evaluation ahead of it, not a settled arrangement.
What the experiment will and will not show
What it can establish
Whether a regulated supply chain is workable: quality control, traceability, taxation and displacement of the criminal supply.
What it cannot
Anything about the wider effects of the tolerance policy itself, which has been running for decades and is not what is being tested.
The scale
Limited by design, to a set of municipalities rather than nationally. That limitation is deliberate and it constrains what can be concluded.
The timeline
Years, with formal evaluation. Nothing about it is quick.
What the policy is often taken to mean
The common assumption
That the Netherlands legalised, and that the arrangement is a model of liberalisation.
What it actually is
A prosecutorial tolerance of retail sale, within a legal framework that never changed, alongside a supply chain that stayed criminal for four decades.
Why the misunderstanding is widespread
Because the visible part — shops operating openly — looks like legalisation, and the invisible part is the whole complication.
What the Dutch position has been
Consistently that this is a pragmatic harm-reduction arrangement rather than an endorsement. That framing has been stable across governments.
The international dimension
The treaty position
The Single Convention on Narcotic Drugs of 1961 and the framework built on it constrain what states parties may do. The tolerance was designed to operate within that.
How it does so
By leaving the prohibition in place. A policy of not prosecuting is legally distinct from authorisation, which is precisely why it took that form.
The pressure it attracted
Sustained criticism from other states and from treaty bodies, particularly over cross-border effects.
Why that shaped the policy
The residence criterion and much of the tightening came from that pressure rather than from domestic concerns.
What none of this has to do with the compliant market
Different legal basis entirely
European compliant products come from listed varieties within the total THC limit, under agricultural and food law. Nothing about the Dutch tolerance applies.
Different material
The compliance limit defines a category that the tolerated sector does not deal in.
Different frameworks
One is a prosecutorial policy on prohibited substances; the other is agricultural regulation of a permitted crop.
Why the association is made anyway
Because the Dutch sector is the most visible cannabis retail in Europe, and visibility attracts association.
What the episode demonstrates
That partial arrangements are unstable
A tolerance covering half a supply chain generated four decades of a problem everyone could name and nobody could fix within the arrangement’s own terms.
That visibility is not legality
The most visible retail sector in Europe operated on prosecutorial discretion throughout, which is a far weaker foundation than it appeared from outside.
That local discretion matters
Most of what has changed came from municipal decisions rather than from national policy. The sector shrank locality by locality.
Why this is worth knowing
Because arrangements elsewhere are frequently described by analogy to this one, and the analogy usually assumes a legalisation that never happened.
What the evaluations found
On market separation
Dutch and international assessments have generally concluded that the policy’s central objective — keeping the two markets apart — was substantially achieved.
On prevalence
Use rates in the Netherlands have consistently sat within the range of comparable European countries rather than above it, which is the finding most often cited in the policy’s favour.
On the supply chain
Uniformly critical. Every serious evaluation identified the criminal upstream as the arrangement’s principal failure, and did so for decades before anything was done.
What the data source is
The European Union Drugs Agency compiles comparable national data across member states, which is what makes the cross-country comparison possible at all.
The cross-border effects
What happened
Towns near the German and Belgian borders received substantial numbers of visitors travelling specifically for access, with associated pressure on local infrastructure and policing.
The municipal response
Border municipalities applied the residence criterion strictly, and several closed premises altogether. It is the clearest case of local discretion reshaping the policy.
The diplomatic dimension
Neighbouring states raised it repeatedly, and much of the tightening from the twenty-tens onward traces to that pressure rather than to domestic politics.
What it demonstrates
A national tolerance in one member state has effects across borders that the state cannot manage alone. It is the structural difficulty with unilateral arrangements in this area.
The word that carries the whole policy
Its literal sense
The Dutch term translates as a policy of condoning or permitting — of looking away from something that remains formally forbidden. It is not a word any legal system uses lightly.
What it concedes
That the thing being permitted is still an offence. The vocabulary is honest about the arrangement in a way that most descriptions of it in other languages are not.
Why the translation matters
Because the English reporting has almost universally rendered it as legalisation, which is the one thing it explicitly is not. The misunderstanding is partly a translation failure.
A short glossary
Tolerance policy
A decision not to prosecute specified offences under stated conditions, leaving the underlying prohibition in place.
The back door problem
The structural contradiction of tolerating retail sale while leaving the supply chain that makes it possible entirely criminal.
The conditions
The criteria under which premises are tolerated: no advertising, nothing from the higher-risk category, no nuisance, no minors, and quantity limits.
The supply chain experiment
A legislated trial establishing licensed production and traceable supply to participating premises in selected municipalities.
Single Convention
The 1961 treaty framework within which the tolerance was designed to operate by leaving the prohibition formally intact.
The five checks that apply to any product
The batch certificate
Request it with the number printed on the packaging. A general document for the product line does not cover this batch, and those differences are exactly what it should record.
Total THC
Delta-9-THC plus THCA multiplied by 0.877, because THCA becomes THC under heat and the limit applies to the sum, not to either value alone.
Price per milligram
A division, not an opinion. The only figure that makes two products comparable, whatever the format, process and pack size.
The claims
Without an authorisation number in the EU Register these are assertions, not data — and for cannabidiol that register contains no authorised entry to this day.
The seller’s details
Company name, address and contact. Without them there is also no counterparty to turn to if something turns out to be wrong.
How this connects to the rest
With the compliant market
Nothing transfers. Different law, different material, different framework entirely.
With the regional production histories
The supply that fed the tolerated sector for decades came partly from the producing regions covered separately.
With regulatory reading generally
The distinction between a tolerance and an authorisation is the same distinction that runs through every claim in this subject.
With the certificate
Nothing here reaches a product document. It is policy history.
What this is for in practice
When the Netherlands is cited as legalised
It is not, and never was. Retail sale is tolerated under conditions and the law is unchanged.
When the arrangement is offered as a model
Note that its central contradiction went unresolved for four decades and is only now being trialled.
When Amsterdam is invoked
The city has been reducing the sector’s visibility for years and says so explicitly.
When comparing to compliant products
There is no comparison. Different legal basis, different material, different regulation.
Frequently asked questions
Did the Netherlands legalise? No. Retail sale is tolerated under prosecutorial guidelines. The underlying prohibition was never removed.
What changed in 1976? The legislation separated substances by assessed risk, with different enforcement approaches, in order to separate the markets.
What are the conditions? No advertising, nothing from the higher-risk category, no nuisance, no minors, and limits on transaction size and stock.
What is the back door problem? Retail sale is tolerated while the supply chain remains entirely criminal. Both are necessary for the premises to operate.
How long did that last? Around four decades. It was identified immediately and only began to be addressed by the recent experiment.
Is the sector growing? No. The number of premises peaked in the nineteen-nineties and has declined steadily since, largely through municipal decisions.
What is Amsterdam’s position? It has been actively reducing the sector’s visibility and discouraging visitors coming specifically for it.
What is the current experiment? A legislated trial establishing licensed production and traceable supply to participating premises in selected municipalities.
Does it affect the compliant market? No. Compliant products operate under agricultural and food law with a content limit. The frameworks do not overlap.
Why is the misunderstanding so common? Because open shopfronts look like legalisation, and the part that was never legal was never visible.
What we check and what we do not
We check what can be checked from documents: official registers, declared contents, certificates of analysis, arithmetic. We test no products, rank nothing, and assert no effects.
The most instructive thing here is a distinction that gets lost in almost every account: a decision not to prosecute is not a decision to permit. Everything that followed — the criminal supply chain, the municipal discretion, the steady shrinking of the sector — follows from that one legal fact.
It is also why the arrangement is a poor model for anything. It resolved half a problem deliberately, lived with the other half for four decades, and is only now running a trial on the part it left out. None of it touches the compliant European market, which is regulated as agriculture and food, and which never needed a tolerance because nothing about it was prohibited.