The cost structure of a licensed cannabis cultivation site
Capex builds the room; opex runs it, and labour and energy usually dominate running costs. A worked 500 m² example shows where the money goes.
On this page
A cultivation licence and a bank loan are really asking the same question in different words: what does this actually cost? Capex is what you spend once to build the room, sunk into concrete, steel and cabling before a single seed goes in. Opex is what you spend every week to keep it running, and it never stops. Confuse the two in a financial model and the numbers look fine right up until the first electricity bill lands.
- Energy + water, one reported upper bound
- 50% of opex
- a single 2020 industry estimate [1]; check current figures for your climate
- Illustrative capex, 500 m² indoor room
- ≈2,000€/m²
- worked example below, not a benchmark
- Illustrative opex, same room
- ≈2,500€/m²/yr
- €1.25m ÷ 500 m², worked example
- Cost/gram, 3 vs 6 harvests a year
- 10.4 vs 5.2€/g
- same room, opex held flat, illustrative
Capex builds the room, opex runs it
Capex, capital expenditure, is the one-off spend that turns an empty shell into a working cultivation site: the room and its segregation, HVAC and dehumidification, lighting, the irrigation and fertigation backbone, environmental controls and sensors, security (CCTV, access control), and the IT and cultivation-record systems a licence requires you to run. You pay it once, then depreciate or amortise it over years.
Opex, operating expenditure, is everything that recurs: labour, energy, nutrients and growing media, packaging, compliance and testing, insurance, and either rent or the debt service on the capex you borrowed to build the room. You pay it every month whether or not the crop is good, and if you underestimate it, a well-built room still fails as a business.
The distinction matters for financing because lenders and investors treat the two differently. Capex is usually financed against the asset itself, a loan or lease secured on the equipment and fit-out, amortised over five to ten years. Opex has to be funded from working capital and, eventually, revenue, and a facility with a beautifully specified capex stack and six weeks of opex runway is a common way for a well-financed grow to run out of cash mid-build. A dehumidifier you buy is capex; the electricity it draws every month for the life of the room is opex. A carbon filter is capex; the media you replace in it every 12 to 18 months is opex.
Two deep-dive pages carry the sourced figures behind this one: capex benchmarks per square metre of canopy and the opex breakdown by category. This page is the skeleton; those are where the numbers live.
Where the build budget goes: the major capex lines
Most licensed indoor builds spend in roughly this order, and the order matters because each line gates the next: you cannot size lighting sensibly before HVAC, and you cannot size HVAC before the shell is segregated and sealed.
Shell and room segregation. Walls, flooring, vapour barriers, doors and airlocks that separate cultivation from packaging, storage and public-facing areas. Licensed facilities usually need this segregation for traceability and pest control even where the building itself is rented, which is why a "cheap" existing warehouse can still need a large fit-out spend.
HVAC and dehumidification. Sized to plant transpiration load, not room volume, this is the line first-time operators most often underbudget. A flowering canopy under high-intensity lighting transpires far more water per hour than the same floor area used for anything else, and undersized dehumidification shows up months later as late-flower botrytis, not as a line item anyone flagged at design time.
Lighting. Fixtures, drivers and the electrical infrastructure to run them. Lighting and HVAC compete for the largest single capex line; which one wins depends on target PPFD, climate, and how much of the cooling load the lighting itself creates.
Irrigation and fertigation backbone. Tanks, dosing pumps, distribution manifolds and drippers or emitters, sized for the room's peak daily water demand, not its average.
Environmental controls and sensors. The controllers, sensors and wiring that hold temperature, humidity and CO2 in range and log the readings a licence typically requires you to keep.
Security. CCTV coverage, access control and, in many jurisdictions, a specific buffer distance or named security role. Conditions here are set by the regulator, not by choice; see regulatory security requirements.
IT and cultivation-record systems. Networking, servers or cloud infrastructure, and the record-keeping software a licence requires for batch tracking and reporting. This library covers what these systems need to do, criteria not vendors; Horus builds one of them, and this page does not recommend it or any other.
Where the running budget goes: the major opex lines
Labour. Cultivation, harvest, trim, sanitation and facility staff, plus the supervision and QA roles a licence typically names explicitly. Indoor cultivation is more labour-intensive than greenhouse or outdoor production at the same canopy area, because climate and irrigation that a greenhouse gets partly free still need managing indoors, and because harvest and trim labour scales with canopy regardless of growing method.
Energy. Lighting, HVAC and dehumidification running around the clock, plus CO2 supplementation where used. This is usually the largest utility-type line and, in high-tariff regions or poorly specified rooms, can rival labour.
Nutrients and media. Fertigation inputs and, for non-recirculating systems, replacement growing media between cycles.
Packaging and consumables. Primary and child-resistant packaging, gloves, sanitiser, and the single-use items a clean-room-style operation goes through continuously.
Compliance, testing and QA. Batch testing at a licensed lab, the QA labour to release or reject batches, and the record-keeping a licensing authority audits. Section 12.1 covers why this line is easy for a first-time operator to underestimate.
Insurance. Property, crop and liability cover; premiums for cannabis cultivation typically run above equivalent horticultural or industrial risk because insurers still treat the sector as higher-risk.
Rent or debt service. Either a lease payment or the interest and principal on the loan that financed the capex; either way, it is fixed regardless of how the crop performs that month.
Fixed costs don't care about your yield
Every opex line above is also either fixed or variable, and the split matters more than the category labels once you start modelling breakeven. Rent or debt service, insurance, and a base staffing floor (the minimum crew needed to run the room safely, even at reduced output) don't move much whether the room is at full canopy utilisation or half of it. Nutrients, growing media, packaging, and the portion of labour that scales with harvest volume do move, roughly in step with how much you actually grow and process.
This is why two facilities with identical annual opex totals can have very different risk profiles. A site with a high fixed-cost share needs to hit a minimum canopy utilisation every single cycle to break even, and a missed harvest or a room down for remediation hurts it far more than it hurts a site built with more of its cost structure variable, for instance through contract labour instead of a large permanent crew. Neither structure is wrong; they trade resilience against a bad month for cost efficiency in a good one, and the right answer depends on how predictable your growing conditions and your market are.
A worked example: costing a 500 m² flower room
The table below is a structural illustration for a hypothetical 500 m² indoor flower room, built to show how the categories above stack up against each other, not a benchmark to plug into your own model. Real figures vary hugely by country, climate, tariff and build quality; use the capex benchmarks and opex breakdown pages linked above for sourced numbers.
A sentence on what to look for: HVAC and dehumidification together outweigh lighting here, and the eight capex lines total to roughly €2,000 per m² of canopy, informed by the structure (not the exact figure) reported for a comparable indoor build [2].
| Capex category | Amount (€) | Share |
|---|---|---|
| HVAC and dehumidification | 320,000 | 32% |
| Lighting | 250,000 | 25% |
| Shell and room segregation | 180,000 | 18% |
| Contingency | 70,000 | 7% |
| Security (CCTV, access control) | 50,000 | 5% |
| Irrigation and fertigation backbone | 60,000 | 6% |
| Environmental controls and sensors | 40,000 | 4% |
| IT and cultivation-software setup | 30,000 | 3% |
| Total | 1,000,000 | 100% |
A sentence on the opex side: labour and energy alone take almost two-thirds of the annual budget, and rent or debt service is the next-largest line even though the room needs no attention to generate it.
| Annual opex category | Amount (€) | Share |
|---|---|---|
| Labour | 480,000 | 38% |
| Energy | 300,000 | 24% |
| Rent / debt service | 200,000 | 16% |
| Nutrients and media | 90,000 | 7% |
| Compliance, testing and QA | 70,000 | 6% |
| Packaging and consumables | 40,000 | 3% |
| Insurance | 35,000 | 3% |
| Maintenance and other | 35,000 | 3% |
| Total | 1,250,000 | 100% |
Why two facilities growing the same crop post wildly different costs per gram
Cost per gram is the number everyone wants and the number this library is most cautious about, because it is a ratio, and both sides of it move independently between two facilities that look identical on paper.
Canopy utilisation. A room designed for five or six harvests a year that actually gets four, because a room turn slipped or a crop failed, still carries almost the same annual opex, divided over less flower. Using the worked room above with annual opex held at €1.25 million and an illustrative 80 g/m² per harvest (500 m² × 80 g = 40 kg per cycle; see yield per square metre indoors for sourced yield ranges rather than this rough assumption), cost per gram on opex alone runs from about €10.42/g at three harvests a year down to about €5.21/g at six, roughly halving as utilisation doubles, with nothing about the growing itself changing:
| Harvests per year | Annual yield (kg) | Cost per gram (opex only) |
|---|---|---|
| 3 | 120 | €10.42 |
| 4 | 160 | €7.81 |
| 5 | 200 | €6.25 |
| 6 | 240 | €5.21 |
In reality a few opex lines, energy and some labour among them, would trim slightly at lower utilisation, so the real swing is somewhat less dramatic than the table suggests; the direction and rough scale hold regardless.
Yield per cycle. Two rooms running the same number of harvests a year still post different numbers if one grows 60 g/m² and the other 100 g/m², whether from cultivar choice, canopy training, or environmental control.
Energy price. A room in a region paying €0.15/kWh and one paying €0.35/kWh can see a two-to-threefold difference in their single largest utility line for identical equipment and identical grams produced; see energy benchmarks for indoor cultivation.
Automation level. Automated fertigation, climate control and trim can shift cost from the labour line to the capex line, lowering opex per gram at high utilisation while raising the fixed cost that has to be covered even when utilisation drops.
For real, sourced cost-per-gram figures rather than this page's illustrative arithmetic, see commercial cost of production benchmarks.
Indoor, greenhouse or outdoor: the same skeleton, different weights
The capex and opex categories above apply to every growing method; what changes is how much weight each one carries.
| Criterion | Indoor | Greenhouse | Outdoor |
|---|---|---|---|
| Relative capex per m² of canopy | Highest | Medium | Lowest |
| Relative energy opex | Highest (full HVAC and lighting load) | Medium (supplemental light and heat, some free light) | Lowest (little or no artificial lighting or climate control) |
| Typical cycles per year, photoperiod cultivar | 5–6 | 3–5 with light deprivation, 1–2 passive | 1, seasonal |
| Climate risk | Lowest, fully controlled | Medium, structure buffers weather | Highest, exposed to weather, pests and theft |
| Verdict | Highest control and highest fixed cost; works in any climate | A middle ground on both capex and energy, with some weather exposure remaining | Lowest capex and energy, but one harvest a year and output at the season's mercy |
Indoor spends the most on climate control and light and gets the most cycles a year in return; outdoor spends the least on both and gets one. Greenhouse sits between the two, and where exactly depends heavily on local climate and whether supplemental lighting is used. The full trade-off, including how climate risk changes the economics, belongs in choosing indoor, greenhouse or outdoor for commercial production.
What a hobbyist's cost model misses at licensed scale
Someone scaling up from a home tent brings a mental model that undercounts two things specifically.
The first is compliance and testing as a real, recurring line, not a one-off registration fee. Batch testing, QA labour to release or reject product, and the record-keeping a regulator audits (see GACP for cultivators and audit readiness) show up every cycle, and unlike most capex, this line grows with the number of batches you run, not just with canopy area.
The second is that canopy sits idle between harvests unless room turns are scheduled deliberately. A home grower's tent is never "down"; a commercial room between strip, clean, and replant can lose days or weeks of canopy time that a hobbyist's mental model has no place for, and that idle time is exactly what the cost-per-gram section above shows driving the biggest single swing in the economics. Getting the room-turn schedule right is worth more to the cost structure than almost any single equipment choice, and it costs nothing to plan.
Once the skeleton here is clear, the next step is putting real, sourced numbers against it for your own market: start with the licensing path if you haven't applied yet, or the capex and opex benchmark pages linked above if you have and are building the model.
Sources
- Smith D, Resource Innovation Institute (2020). A resource-efficient cannabis industry starts with benchmarking. Health Europa Accessed 2026-09-26.
- Northstar Financial Advisory (2024). Cost to start a cannabis grow operation: 2025 capital breakdown Accessed 2026-09-26.
- Resource Innovation Institute (n.d.). Cannabis PowerScore Accessed 2026-09-26.