Germany Medical Cannabis Market Report, Size, Share and Forecast 2026–2033

Germany imported 201.1 tonnes of cannabis for medical and scientific purposes in 2025 — 6.19 times the 32.5 tonnes of 2023, 2.76 times the 72.9 tonnes of 2024, and about 551 kilograms a day. Dispensed through pharmacies at the prices patients actually paid, that is a market of EUR 988.5 million, or USD 1,106.4 million at the ECB reference rate of 1.1193 for 23 October 2026. Everything else in this report explains why a mass that tripled and a value that did not are not in conflict — and which of the two is about to change.

Germany is the largest regulated medical-cannabis market in the world by volume, and it became one in under two years. The mechanism was legislative rather than clinical. On 1 April 2024 the Medizinal-Cannabisgesetz took cannabis for medical purposes out of the narcotics schedules and made it an ordinary prescription-only medicine. A physician no longer needed a narcotics prescription, and a patient no longer needed a statutory insurer's approval to be treated. What had been a narrow, heavily gated therapy became a standard outpatient prescription — still off-label, still without a marketing authorisation, but administratively ordinary.

Two comparisons explain everything that followed, and both come from the ministry that now wants to reverse part of it. Between the first and second halves of 2024, imports of medical cannabis flower rose 170 per cent while prescriptions charged to statutory health insurance rose 9 per cent. In the first half of 2025, imports reached roughly 80 tonnes against roughly 19 tonnes a year earlier — the ministry describes that as an increase of more than 400 per cent, and on its own rounded figures the arithmetic is +321.1 per cent — while statutory prescriptions again rose by single digits. The ministry's own FAQ draws the conclusion the industry has been reluctant to state: the import surge is the supply of an increasing number of self-payers holding private prescriptions outside the statutory system, ordered through telemedicine platforms, in many cases without any personal contact between the patient and either a physician or a pharmacist.

A market built on a prescription form, not on a patient cohort

This is the single most important structural fact about German medical cannabis, and it is a fact about administration rather than about medicine. The growth mechanism and the corrective instrument are the same object: the prescription. When the state decided that a standard prescription was sufficient, it created a market whose demand could be generated by advertising and fulfilled by post. When it decided that a personal consultation was required and that flowers could no longer be mailed, it proposed to remove both of those properties at once.

The volume is not small and it is not marginal. 201.1 tonnes is roughly half a kilogram per minute, every minute, for a year. A market of that size arriving in 24 months is not a gradual diffusion of therapy; it is a distribution channel being switched on. The relevant question for anyone sizing this market is therefore not how many Germans have a condition cannabis treats — that number moves slowly and is bounded by epidemiology — but how long the channel stays open. We treat those as separate questions throughout this report, and the forecast in it is essentially a view on the second.

Price: the number that makes the tonnage misleading

In January 2025 the average price patients paid for a gram of medical cannabis flower in Germany was EUR 8.33. In December it was EUR 5.23 — a fall of 37.2 per cent in eleven months, recorded in the operator survey published as the Cannabis Barometer. Over the same period the share of flower selling below EUR 6 per gram rose from a small fraction of the range to roughly four-fifths of it. The compression is not a promotional episode; it is what happens when the number of licensed exporters and the number of dispensed SKUs both rise faster than the number of patients, and when the patient is paying out of pocket and therefore shopping on price.

A market in which mass rose 175.86 per cent and unit price fell 37.2 per cent is a market whose revenue grew by far less than its tonnage suggests. That asymmetry is the reason this report publishes a value forecast of 4.86 per cent for a market that has been growing at triple digits — and the reason the tonnage figure, taken alone, has misled almost everyone who has written about it.

What this report measures, and the three boundaries

We size the market from the mass that reaches patients, multiplied by the realised price per gram at the pharmacy. The input is the volume authorised for import by the Federal Institute for Drugs and Medical Devices (BfArM), which publishes a quarterly series; the price input is the observed patient-paid average for flower. Three boundaries follow, and a reader should hold all three.

First, authorised imports are not dispensed grams. Between the port and the patient sit processing losses, re-export, regulatory clearance lag and inventory build. We apply a 65 to 80 per cent sell-through band, and we state plainly that this band is our assumption, not an observed statistic — no German authority publishes the share of imported mass that is dispensed. Our point estimate uses the midpoint of 72.5 per cent, which gives 145.8 tonnes dispensed in 2025.

Second, the price is a retail average, not a reimbursement rate. The EUR 5.23 to EUR 8.33 range is what self-paying patients were charged. Where statutory insurance covers the cost, pharmacies may charge materially more — the same survey cites a ceiling of EUR 19.04 per gram for covered prescriptions. A market whose value is quoted from reimbursement data and a market whose value is quoted from patient-paid prices are therefore not comparable, and the gap between them is not a margin.

Third, the value is a pharmacy-level figure. It is not the CIF value of imports and it is not manufacturer revenue. It is what the patient or the insurer paid, which is the only measure that can be compared with a published price per gram.

Market size, 2025

On those inputs, Germany's medical-cannabis market was worth EUR 988.5 million in 2025. The sell-through band alone takes it to EUR 886.3 million to EUR 1,090.8 million. Widening to the price extremes observed during the year — the January price against the December price — gives EUR 683.7 million to EUR 1,340.1 million. That second, wider band is the honest range for this market: the price fell so far inside a single calendar year that the choice of which month's price to apply moves the answer by more than the choice of sell-through does.

The value that does not reconcile

A trade association figure widely quoted for this market is approximately EUR 2 billion for 2025, roughly double 2024. We report it because it is in circulation and a reader will encounter it, and we do not use it, because it cannot be reconciled with either the volume or the price series.

The arithmetic is simple enough to state. Two billion euros spread across 201.1 tonnes of imports implies EUR 9.95 per gram on the impossible assumption that every imported gram reached a patient — a price above the highest monthly average observed at any point in 2025. Spread across the 145.8 tonnes we believe were actually dispensed, it implies EUR 13.72 per gram, which is roughly two and a half times what patients paid in December. Alternatively, at December's price, EUR 2 billion would require 382.4 tonnes to be dispensed — nearly twice the total volume that entered the country. Even the theoretical ceiling of the import record at the lowest observed monthly price is EUR 1,051.8 million.

Our reading is that the EUR 2 billion figure is not a measurement of the same thing we are measuring. It is consistent with a reimbursement-rate basis, where covered prescriptions are charged at up to EUR 19.04 per gram, or with volumes outside the BfArM import series. Either is a legitimate subject; neither is a patient-paid sell-out figure, and mixing them produces a market size that no operator can reproduce from its own revenue.

The forecast to 2033: kilos, not euros

We build the forecast from its two drivers rather than by asserting a growth rate, because in this market the two move in opposite directions and a single rate hides that. Dispensed mass grows; realised price per gram falls. The product of the two is the forecast, and it is materially flatter than either component.

The base case puts dispensed mass at 250.5 tonnes by 2033, growing 7.0 per cent a year — a sharp deceleration from 2025 but still expansion, consistent with a market that has established a self-payer channel that survives in some form. It puts the realised price at EUR 5.77 per gram, falling 2.0 per cent a year, which is a fraction of the compression actually observed in 2025. The product is EUR 1,445.4 million, a CAGR of 4.86 per cent.

The conservative case is the amendment passing in a form close to the current draft and price decay continuing at 3.5 per cent a year: EUR 1,017.4 million by 2033, a CAGR of 0.36 per cent over eight years. The optimistic case is a channel that reopens and a patient base that keeps compounding at 11 per cent a year with only 1 per cent annual price decay: EUR 2,103.4 million, a CAGR of 9.90 per cent.

The band is wider than we normally publish, and the reason is structural rather than an excess of caution: a single-rate forecast compounds one number, while this one multiplies two, each of which is moving by several per cent a year in opposite directions. The sensitivity is worth stating in one line — one percentage point of value CAGR is worth EUR 384.2 million of 2033 revenue, more than a third of the base-case market. Anyone using this forecast to underwrite capacity should hold the band, not the point.

How the report is organised

The chapters that follow take the market apart in a fixed order. The competitive landscape looks at where the grams actually come from — Canada first, at 45.97 per cent of 2024 import volume, with the top three origins at 80.1 per cent — and at why a market that keeps adding origins still concentrates. The value chain follows a shipment from an EU-GMP facility in Ontario or Portugal to a pharmacy counter, and shows where the 55.3 tonnes that never reach a patient go. Technology and innovation covers the one place where German medical cannabis is genuinely productive: cultivation, genetics and formulation for a self-payer who reads a price list. Investment activity considers what capital has been buying, which in this market has largely been distribution rather than production. Market risk sets out the six exposures, of which the amendment is only the largest. The regulatory landscape sets out the current law and the proposed law side by side, including the European-law questions that the Bundestag's own scientific service has raised against the mail-order prohibition.

Readers coming to this market from the pharmaceutical side should read this report alongside the global cannabis market and global prescription drugs reports, because Germany is where a consumer-adjacent category was temporarily classified as ordinary medicine and is now being reclassified back. Readers from the reimbursement side should read it alongside private health insurance, because the distinction between a self-payer and an insured patient is doing more work in this market than any clinical variable.

Method note: import volumes are BfArM's published quarterly series for cannabis for medical and medical-scientific purposes, summable to 32.5 tonnes (2023), 72.9 tonnes (2024) and 201.1 tonnes (2025); quarters are revised after first publication, so the latest full year may move. Prices are the observed patient-paid average for flower from the Cannabis Barometer series and are survey-based. The sell-through band, the price path and the resulting value are Pheonix's own construction, disclosed here so that a reader can substitute their own assumption and see what it does. No syndicated market estimate is used or cited. Currency conversion uses the ECB reference rate of 1.1193 for 23 October 2026.

Table des matières

1. Executive summary
2. A market built on a prescription form, not on a patient cohort
3. Price: the number that makes the tonnage misleading
4. What this report measures, and the three boundaries
5. Market size, 2025
6. The value that does not reconcile
7. The forecast to 2033: kilos, not euros
8. Competitive landscape
9. Value chain
10. Technology and innovation
11. Investment activity
12. Market risk
13. Regulatory landscape
14. Frequently asked questions

Paysage concurrentiel

Where the grams come from

Germany publishes its import origins annually, and 2024 is the last complete year of that table: 72,122 kilograms, in a list that is short at the top and long at the bottom.

Canada at 45.97 per cent and Portugal at 23.89 per cent are not two versions of the same fact. Canada is an origin: it has the cheapest large-scale EU-GMP production in the world for this product and its position is a cost-of-production position. Portugal is a manufacturing jurisdiction — it hosts EU-GMP cannabis cultivation and processing built for export and it sits inside the single market — and its position is a capacity position, which makes it the most contestable entry in the table, because any member state that certifies the same facilities can compete for it. Denmark at 10.25 per cent is the proof of that: a licensed cultivation industry built specifically for pharmaceutical export, holding double-digit share against Canadian cost structures on the strength of proximity and qualification.

Then the tail. North Macedonia at 2,666 kilograms, Spain at 2,212, the Netherlands at 2,083, Uruguay at 1,687, Australia at 1,428 — and 4,265 kilograms spread across every remaining origin combined. A market that imported 201.1 tonnes in 2025 was still sourcing the overwhelming majority of it from three countries.

Why a market that tripled in volume still concentrates

The mechanism is the import authorisation. A BfArM authorisation is granted for a named exporter, a named importer, a stated quantity and a defined consignment, and the exporter’s facility must hold EU-GMP certification for that product form. Qualifying a new origin is therefore not a decision, it is a file: a certificate to be reviewed, a batch to be tested, a wholesale authorisation to be matched to it. An importer that already has three qualified corridors has no reason to qualify a fourth in a year when the price of a gram is falling 37.2 per cent. Growth flows to the corridors that already exist, which is exactly what the table shows, and it is why the 2025 surge of 175.86 per cent went to the incumbents.

This is the structural difference between medical cannabis and a registration-protected input market. There, the barrier is a domestic administrative file that protects a local price. Here the barrier is a foreign manufacturer’s certification plus a shipment-level permit, and it protects nothing except the relationship. It is far easier to enter than a registration regime and far easier to lose: a supplier that fails an audit or misses a delivery window is replaced by a qualified alternative within a season, not within a filing cycle.

The tail grows when price decides

The pattern the annual table does not yet show is diversification at the margin. When the patient pays out of pocket and the average gram is EUR 5.23, the origin that wins the next order is the one with the lowest landed cost — and the cheapest competent supplier is no longer necessarily Canadian. Latin American and Southern African origins are structurally cheaper producers at the same EU-GMP specification; what constrains them is qualification, not cost. Every point of share they take comes out of the top three, not out of the market.

The evidence that this is already happening sits on the pharmacy shelf rather than in the trade table: the number of distinct flower products available to German patients rose from 468 to 724 during 2025 in Bloomwell’s analysis of the market. More origins, more products, lower average price — the three move together, and none of them is a sign of a market becoming more profitable.

Three layers of competition, and only one of them sets price

Exporters. EU-GMP cultivators and manufacturers: the Canadian licensed producers — Canopy Growth among the largest — alongside Danish, Portuguese and Macedonian facilities and a growing number of Colombian and Southern African operations. They compete on cost per gram at a given specification, and on reliability of supply into a market where the unit price is falling.

Importers and wholesalers. The layer that holds the import authorisation and the wholesale licence, takes title to the goods and carries the inventory. This is where the commercial risk of the German market actually sits, and it is the layer with the least control over price.

Prescribers and dispensing pharmacies, increasingly intermediated by telemedicine platforms. The layer that generates the prescription. It does not take title to a gram of inventory, and it is paid per consultation and per dispensing event — which gives it an incentive to maximise the number of patients, not to defend the price of the product.

That split explains the apparent paradox of this market: intense competition at the supplier end, a collapsing unit price, and a demand-generating layer that benefits from the collapse because cheaper grams bring more patients. We classify competitive intensity as high and the structure as fragmented, with the qualification that the fragmentation sits at the top of the chain rather than the bottom. There are many more origins than there are authorised German importers.

What consolidation would actually look like

Not a merger of growers. The asset that could consolidate is the importer’s authorisation together with its pharmacy relationships, and the reason it has not happened is that both are cheap to replicate and hard to value: an import authorisation is not transferable in the way a registration is, and a pharmacy relationship in a mail-order market is a transaction rather than a relationship. Consolidation in this market, when it comes, will be a consolidation of patient acquisition rather than of supply.

Chaîne de valeur

The chain in six steps

Cultivation and manufacture outside Germany, in an EU-GMP facility. Export authorisation at origin, granted against the German authorisation. Import authorisation from the BfArM under section 4 MedCanG and the narcotics foreign-trade ordinance, granted per consignment. Import and wholesale: an authorised importer takes title and holds the goods. Dispensing: a pharmacy dispenses against a prescription, with a counselling obligation. The patient, or the statutory insurer, pays.

Two features of that chain decide the economics of everything above it.

First: the chain is authorised twice, and neither authorisation is a market position

Every consignment needs an export authorisation at origin and an import authorisation in Germany, and the German authorisation is quantity-specific. This is why the market ran into a national ceiling in 2025 — and the ceiling is not a tariff or a trade quota. It is the estimate of annual requirement that Germany files with the International Narcotics Control Board, which the authorisation system enforces shipment by shipment. When the original allowance of 122 tonnes was exhausted in mid-September 2025, imports stopped, and they resumed only after the BfArM raised the maximum to approximately 192.5 tonnes. A market whose physical limit is an administrative number can be halted in a week and restarted in a week, and 2025 is the year that stopped being theoretical.

Note what the 2025 outcome means for anyone modelling this market: the volume that arrived, 201.1 tonnes, exceeded the raised ceiling of 192.5 tonnes. The quota did not bind in the end. The uncertainty it created did.

Second: 55.3 tonnes of the 2025 import never reached a patient

On our point estimate, 145.8 of the 201.1 tonnes imported in 2025 were dispensed and 55.3 tonnes were not. That gap is not waste in the ordinary sense. It contains processing losses — flower converted to extract loses mass — re-export to other European markets, product held as inventory against future prescriptions, and batches awaiting release.

It is also where a falling price does the most damage, because inventory bought at EUR 8.33 a gram and sold at EUR 5.23 a gram is a loss that no volume growth recovers. This is the central commercial fact of the chain: the importer takes title, the patient sets the price, and the price fell 37.2 per cent inside the calendar year. A wholesaler that bought correctly in January and was still holding in December lost more than the margin it earned, and it did so while its tonnage was rising. The 175.86 per cent volume growth that reads as a boom in every headline is, at the level of the balance sheet that carries the inventory, a mark-to-market event.

The layer that generates demand does not own the inventory

The telemedicine platforms that assemble the prescriptions sit outside the physical chain. They are paid per consultation and per dispensed prescription; they do not take title to a gram and they do not carry the mark-to-market. The structural consequence is asymmetric: the party with the strongest incentive to grow demand has no incentive to defend price, and the party holding the inventory has no control over demand.

We state that as an inference from the structure and from the published price series, not as a verified margin analysis. No German operator publishes platform-level unit economics that we can cite, and we will not construct them.

Domestic cultivation and the cost floor

Germany licensed domestic cultivation under the MedCanG, and the volumes remain small against imports: in its submission to the INCB, the BfArM projects roughly 10 tonnes of domestic production for 2026, against 201.1 tonnes imported in 2025 — a projection reported from the BfArM’s own filing.

The domestic industry’s problem is not quality; German flower is well regarded. It is the cost floor. A German glasshouse carrying German labour, energy and certification costs cannot undercut a Canadian or Portuguese facility selling at EUR 5.23 a gram without a price premium that the self-paying patient has demonstrably refused to pay. Domestic cultivation in this market is a security-of-supply argument, not a cost argument, and it will be sized by policy rather than by demand.

Where the chain breaks

There is exactly one fragile link and everyone in the market knows which one it is: the last. If flower can no longer be delivered by mail and must be dispensed in a presence pharmacy after a personal medical consultation, the number of patients a platform can serve in a day falls from an internet-scale number to a physical one. Nothing upstream changes — the exporters, the authorisations, the wholesale licences and the inventory all remain in place — which is precisely why the volume reaction would be fast and the capacity reaction would be slow. An exporter that has qualified a corridor and built for German volume cannot redirect it in a quarter; a platform that has built its business on mail order cannot replace the channel with a shop network in one either.

Activité d'investissement

What capital is actually buying here

In most pharmaceutical markets capital buys molecules and the rights to sell them. In this one it buys a corridor and the right to import along it, or it buys a prescription flow. Almost none of it buys science, because there is no proprietary molecule to buy: cannabis flower is dispensed off-label, without a marketing authorisation, so the intellectual property that would normally anchor an investment does not exist in this category.

That single fact sorts every plausible investment in the German market into three types, and it ranks them.

Patient acquisition. The asset that captured the growth of 2024–25 was the ability to generate prescriptions — a platform, a brand, a funnel, a dispensing pharmacy behind it. It requires working capital rather than capex, it scales without a factory, and it is the only layer of the chain whose revenue grew faster than its cost of goods during the year prices fell. It is also the layer most exposed to the amendment, because its entire proposition is the remote prescription.

Import and wholesale capacity. An authorisation, a wholesale licence, a validated cold-and-dry logistics chain and the working capital to hold inventory. This is a real asset with a real return, and 2025 was the year it became an uncomfortable one: the importer bought at January prices and sold at December prices. The volume tripled and the inventory lost value while it sat.

Cultivation. The most capital-intensive and the least defensible of the three. A German cultivation asset competes against Canadian cost of production at a EUR 5.23 retail gram with German labour, German energy and German certification costs. It is justified by security of supply and by policy, not by unit economics, and it should be underwritten as infrastructure rather than as a growth business.

Why the disclosed deal record is thin, and what that means

We found no transaction in the German medical-cannabis channel that we can cite with a verified value, and we will not construct a deal count from press summaries. That absence is itself informative, and it is a structural observation rather than a claim that no deal has ever happened: the assets that matter here are authorisations and patient flows, neither of which is a clean transferable thing. An import authorisation attaches to a named importer and a defined consignment; a patient flow attaches to a brand and a prescriber relationship that can migrate in a fortnight. Buyers price uncertainty, and this market supplies a great deal of it.

We therefore classify the investment trend as rising on operational evidence — volume, corridor count, product count — rather than on disclosed transactions, and we flag that the same evidence supports a much weaker reading of enterprise value, because the growth was bought with inventory that depreciated.

The 2025 lesson, in one event

The most consequential event of 2025 for an investor was not a funding round. It was that imports stopped in mid-September because a national allowance filed with the International Narcotics Control Board had been used up, and resumed only after the BfArM raised the ceiling to approximately 192.5 tonnes. Any capacity built on the assumption that supply is a commercial variable is mis-specified: in this market supply is an administrative permission, renewed annually, and the entity that grants it can exhaust it without warning and without a consultation period.

The rational response is to hold optionality rather than capacity — multiple qualified origins, a wholesale licence that is not tied to one corridor, and a customer base that does not depend on a single prescribing channel. That is the opposite of the consolidation logic that usually follows a tripling of volume, and it is why this market has grown without concentrating.

The growth case, and its ceiling

The bull case is simple and it was right for two years: a large European population, a prescription requirement that had just become ordinary, and a product whose cost of production was falling. The ceiling on it is not epidemiological. It is that value equals mass multiplied by price, and in this market mass and price have moved in opposite directions by large amounts — 175.86 per cent up, 37.2 per cent down, in the same year. An investor underwriting volume growth in this category is underwriting a revenue line that has historically grown at a fraction of it, and the eight-year sensitivity is stated plainly in the market-size chapter: one percentage point of value CAGR is worth EUR 384.2 million of 2033 revenue.

Technologie et innovation

Germany’s innovation in this market is workflow, not discovery

The most important technical change in German medical cannabis was not a cultivar, an extraction method or a delivery device. It was the removal, on 1 April 2024, of the narcotics prescription. Overnight a therapy that required a special secure prescription document became one that any physician could write on a standard form. Nothing about the plant changed. Everything about the friction changed, and the friction was the market.

That is the pattern to hold on to when assessing innovation in this category. The chemistry is not novel — cannabis flower has been dispensed off-label for a decade in this market — and the clinical evidence base has not materially moved. What has moved is the cost and speed of getting a prescription to a patient, and the unit cost of moving a gram from a validated facility to a pharmacy counter. Both are logistics and software problems, and both are where the German industry has actually spent its engineering effort.

Three places where the technology is real

Cultivation and consistency. The product that wins a price-led market is not the highest-THC flower; it is the flower whose specified profile arrives identical batch after batch, because a pharmacy cannot substitute and a patient will notice. Consistency is a cultivation-control problem — environmental control, genetic stability, pest management without banned actives, post-harvest handling — and it is where German and Danish producers compete with Canadian ones on something other than cost.

Formulation away from flower. Extracts, oils and cannabinoid preparations are less exposed to the price of a gram of flower and more exposed to pharmaceutical manufacturing capability. They are also the part of the category where a producer can hold a specification that a competitor cannot copy overnight. Where a producer wants to escape the commodity, this is the door — and it is the door the current policy debate does not close, because the amendment restricts flower specifically.

The prescription and dispensing layer. Asynchronous consultation, prescription generation, pharmacy routing, stock visibility and delivery tracking. This is where the market’s real innovation has happened and where its regulatory risk is concentrated, because the same systems that made the channel efficient are the ones the amendment is designed to bypass.

What innovation cannot fix: there is no marketing authorisation

Cannabis flower in Germany is dispensed in off-label use. There is no approved product with a reviewed label, a defined indication set and a clinical dossier behind it, and that single missing artefact caps what innovation in this category can mean. It caps the clinical claim, it caps reimbursement, and it caps the price a producer can defend, because a product without a label competes purely on specification and cost.

It also explains why the category’s product-development signal looks the way it does: the number of distinct flower products rose from 468 to 724 during 2025. That is a very large number of variants and a very small amount of clinical differentiation, and it is the signature of a market innovating on SKU count rather than on evidence.

Innovation intensity, honestly assessed

We classify technology maturity as growth, innovation intensity as moderate and patent activity as low. The moderate rating is for the manufacturing and formulation layer, where EU-GMP process development is genuine and where the industry will have to invest if it wants to survive a EUR 5.23 gram. The low patent rating is not a criticism of the companies; it is a statement that the underlying molecule is prior art and that the protectable space in this category is process and brand rather than composition.

Anyone expecting this market to innovate its way out of price deflation is expecting the wrong thing. The technology in German medical cannabis has already done its work — it made the channel frictionless — and the constraint on the market now is a statute, not a specification.

Risque de marché

Six exposures, and only one of them is priced

1. The regulatory reversal is the market

This is not one risk among six; it is the risk that contains the others. The ministry’s own analysis states that the import surge is attributable to self-payers with private prescriptions obtained outside the statutory system, largely through telemedicine platforms, and its proposed remedy is to require a personal consultation for flower and to prohibit delivery by mail. If that passes in close to its current form, the mechanism that produced a 6.19-fold increase in three years is removed by statute, and the suppliers, importers and inventory that were built around it become capacity without a channel.

Two things make this harder to price than a normal policy risk. First, the bill has sat in the health committee since December 2025 without resolution, so neither outcome is the base case by default. Second, the current rules are already being enforced through general law: courts have found that a merely written medical history is insufficient for a cannabis prescription, and platform advertising has been treated as prohibited public advertising. A market can therefore tighten without the amendment passing at all.

2. The inventory position is a short position on price

An importer buys a consignment at a landed cost, holds it through release and distribution, and sells it at a price set by a patient who is paying cash and comparing products. When that price falls 37.2 per cent in a year, every tonne held through the fall is a loss, and rising volume makes the loss bigger rather than smaller. The tonnage that reads as growth in the trade statistics is, in the balance sheet of the party carrying it, a mark-to-market event. That is why the value of this market can fall while its volume rises, and why a forecast built on tonnage alone will be wrong in a specific and predictable direction.

3. Demand is generated by one channel

The layer that creates the prescription is narrow: a set of telemedicine platforms and the pharmacies that dispense behind them. Concentration in demand generation is more dangerous than concentration in supply, because supply can be re-qualified in a season and a patient channel cannot be rebuilt in a quarter. It is also the exposure that most analysts miss, because the trade statistics list many origins and read as diversification when they are in fact diversification of the same single channel’s supply.

4. The quota is a physical limit set by an administrative act

Germany’s licit supply is bounded by the estimate of requirement it files annually with the International Narcotics Control Board. In 2025 that bound was reached: imports stopped in mid-September when the original allowance of 122 tonnes was exhausted, and resumed after the ceiling was raised to approximately 192.5 tonnes. The volume that ultimately arrived, 201.1 tonnes, exceeded even the raised figure. The risk is not that the ceiling is low — it is that it is discretionary, revised without a consultation period, and enforced consignment by consignment.

5. Three countries supply four-fifths of the market

Canada at 45.97 per cent and the top three at 80.1 per cent of 2024 volume is a moderate exposure rather than a high one, because the product is fungible across qualified origins and there are more willing exporters than there are qualified corridors. A disruption to one origin is a re-qualification problem, not a supply crisis. The exposure that matters is regulatory, not logistical: if a major origin lost its EU-GMP standing, the replacement would take longer to qualify than the market holds inventory for.

6. The reimbursement question nobody is asking yet

Statutory reimbursement for cannabis flower is small in this market — the whole point of the ministry’s analysis is that the growth is not insured patients. That makes the reimbursement file look irrelevant. It becomes relevant in one scenario: if the self-payer channel is restricted, the only remaining demand is insured demand, and a separate 2026 draft would remove flower from statutory reimbursement altogether. The two proposals, taken together, would leave flower as a privately paid prescription available only after a personal consultation and only by collection from a pharmacy. That is a coherent policy position. It is also a market roughly the size of the conservative case in this report.

What would falsify the forecast

The base case of EUR 1,445.4 million by 2033 fails in both directions and the triggers are observable. Downside: the amendment passes with the mail-order prohibition intact and dispensed mass grows at 4 per cent rather than 7 per cent, which is the EUR 1,017.4 million case. Upside: the amendment fails or is substantially amended in the Bundesrat, the channel remains open, and price decay slows to 1 per cent a year — the EUR 2,103.4 million case. The single indicator to watch is the quarterly BfArM import series: a first sequential decline has already appeared, with Q4 2025 at 56,625 kilograms against Q3’s 59,076, a fall of 4.15 per cent. Two more quarters in that direction would confirm the conservative path before any statute is enacted.

Paysage réglementaire

The law as it stands

Medical cannabis in Germany is governed by the Medizinal-Cannabisgesetz, in force since 1 April 2024. The Act did two things that matter commercially. It removed cannabis for medical purposes from the schedules of the narcotics law, so that it is no longer a narcotic drug but a prescription-only medicine; and it created a licensing regime for handling it, under which any import, export or transit requires an authorisation from the Federal Institute for Drugs and Medical Devices under section 4 of the Act, alongside the requirements of the narcotics foreign-trade ordinance.

Three consequences follow. A physician may prescribe cannabis flower or extracts on an ordinary prescription, without a narcotics prescription form and without the prior approval of a statutory insurer. A patient may be treated as a self-payer, and a large and growing share now is. And the state’s control has moved from the prescription to the shipment: every consignment needs a permit, and the total volume is bounded by the estimate of annual requirement Germany files with the International Narcotics Control Board. That is why the 2025 quota exhaustion was possible in a market with no shortage of suppliers.

Flower remains without a marketing authorisation. It is dispensed in off-label use, and statutory reimbursement is available only within the limits of section 31(6) of the Social Code Book V. The absence of an approved product is the reason this market has no proprietary molecule, no clinical dossier and no price that a label can defend.

The law being proposed

The draft First Act Amending the Medizinal-Cannabisgesetz — Bundestag Drucksache 21/3061 — would change three things, each aimed at the same target.

A personal consultation. Cannabis flower could be prescribed only after personal contact between patient and physician, in the practice or on a home visit. Follow-up prescriptions could be issued remotely, but only where a personal contact has taken place within the previous four quarters.

No mail delivery. Flower could be dispensed to patients only in a presence pharmacy against the prescription. Delivery by courier service remains lawful; delivery by mail does not.

No purely remote treatment. The combination of an online questionnaire, a remote prescriber and a mail-order pharmacy — the model that produced the volume — would no longer be a lawful route to a gram of flower.

The ministry’s stated rationale is not fiscal. It is that imports rose more than 400 per cent in the first half of 2025 while statutory prescriptions rose in single digits, that platforms had established a route by which patients obtained flower without any personal contact with a physician or a pharmacist, and that a medicine carrying a dependence risk was being supplied in that way without a reviewed label. The government’s position, put plainly in the Bundestag debate, is that it does not believe Germany acquired 400 per cent more pain patients in a year.

Where the bill stands

The draft has been before the health committee since December 2025 and had not been enacted at the time of writing. It has attracted substantial legal criticism on European-law grounds: the Bundestag’s own European affairs service examined whether the prohibition of telemedical prescribing and of mail delivery is compatible with the free movement of goods and services, and practitioners have made the same argument — noting the asymmetry that the same household may lawfully hold three cannabis plants and 50 grams of home-grown cannabis under the consumer law while a pharmacist-controlled mail delivery of a prescribed medicine would be prohibited.

Separately, a further draft dated April 2026 would remove cannabis flower from statutory reimbursement altogether. It has not advanced either. The legislative position is therefore best described as two live proposals, neither enacted, both pointed at the same channel from different directions.

What is already being enforced without a new statute

The most under-appreciated feature of this regulatory landscape is that tightening does not require the amendment. General medical and advertising law is already being applied to the telemedicine model: a Bavarian court held in June 2025 that a merely written medical history is not sufficient for a cannabis prescription, and the Federal Court of Justice held in March 2026 that a platform’s advertising for cannabis treatment was unlawful public advertising for a prescription-only medicine. Both are citable reasons why a platform may have to change its model before any new Act is passed.

For a supplier, the practical implication is that the risk is continuous rather than binary. The market can tighten through litigation, through the quota, or through the Act, and each of those arrives on a different timetable.

The mismatch, stated plainly

German law asks one question about a gram of medical cannabis: was it prescribed by a physician who had the information and the contact required to prescribe it responsibly? The market asks a different one: how efficiently can a prescription be produced for a patient who is willing to pay? Those are not the same question, and the distance between them is where this market’s regulatory risk lives.

Nothing in the current architecture closes that distance automatically. The Act proposes to close it by requiring personal contact, which will work, and by prohibiting mail delivery, which will also work — and both will reduce the volume. The alternative that nobody has legislated is the one the market itself could supply: an approved product with a label, an indicated set of uses and a clinical dossier, which would move the question from how the prescription was generated to whether the medicine works. Until that exists, this market will keep being governed by its distribution rules rather than by its evidence.