Illinois
Cannabis Testing Data
Every completed compliance test in Illinois's regulated cannabis market, straight from the state's own records. 100,647 packages across 42 named operators and 6 named labs. Sep 2021 – Feb 2026.
Source: Illinois cannabis lab-testing records via public-records (FOIA) request, tagged to license class against the state's cultivation-center and craft-grower registries. Data through February 2026 · built June 2026.
Illinois opened the door, then built the room too small to stand up in.
Illinois converted its 21 medical-program cultivation centers into the adult-use incumbent tier, then issued 86 craft licenses to open the market up. Those 21 licenses are held by just 18 operators — Cresco Labs runs three sites, Green Thumb runs two — so the incumbent tier is even more concentrated than the license count looks. The craft license came with a hard limit the incumbents never faced: a canopy cap that started at 5,000 square feet, which Illinois raised to 14,000 in January 2024 with applications opening that March. Five years of the state's own testing data shows what that structure produced — and how little the expansion has moved it yet.
cultivation-center licenses converted from the medical program — the entire incumbent tier, held by 18 operators, and every site produces tested flower.
of tested flower still came from those 21 incumbent sites across all of 2025 — craft growers, 11.8%.
craft licenses ever appear in the flower testing record — fewer than 1 in 4.
statutory canopy advantage at base — 210,000 sq ft per incumbent against 5,000 for a craft grower, and never better than 15:1 even at the 14,000 sq ft maximum a craft grower can reach.
The funnel
Start with the funnel. Illinois issued 86 craft-grower licenses. 31 carry IDOA's "Operational Approved" flag — a registry field the agency does not formally define. 20 ever appear in the flower testing record — with many still building, and others growing for extraction and manufacturing channels a flower count never sees. The headline number sounds like a new tier of competitors. The record shows 20 growers who made it far enough to test a package, against an incumbent tier of 21 sites — held by 18 operators — that has produced tested flower from the first quarter of the data forward. This is not a story about craft growers who couldn't compete. It is a story about a license structure that decided who could compete before anyone planted — and a buildout clock that means the tier this data captures is still arriving.
The craft expansion's applications only opened in March 2024. Even after a craft grower breaks ground, operational sign-off runs the better part of a year — on top of the multi-year licensing delay before they could build at all, and against incumbents whose cultivation-scale facilities took 12 to 24 months and millions to stand up in the first place. A tier that started building this recently has not finished arriving.
The cap
The mechanism is canopy. A cultivation center may flower up to 210,000 square feet. A craft grower starts at 5,000 — one-forty-second the space — and may expand to 14,000 only after clearing an operational waiting period, a construction approval, fees, and the capital to build three times the room. That's the ceiling. Fully expanded, at the absolute legal maximum, a craft grower is still entitled to one-fifteenth of an incumbent's flowering canopy. The structure never lets craft approach parity. It was not designed to. And the gap the data actually measures — incumbents averaging 4,487 tested packages per operator against 220 for craft, roughly twenty to one — sits inside the band the statute draws. The data does not exaggerate the cap. It lands where the law said it would.
The finding is volume, not quality.
Read the gap correctly. The incumbents average 4,487 tested packages per operator; the craft growers who reach market average 220 — roughly twenty to one. That is the whole story, and it is a story about size, not standards. The structure caps how much canopy a craft grower can put under light. It says nothing about who grows clean flower. What the data measures is a tier held small by design — not a tier that couldn't keep up.
Who produces the flower Illinois tests, by quarter
Full quarters only; partial periods excluded. Cultivation centers produced 100% of tested flower through 2022, and still produced the overwhelming majority of it at the start of 2025. This is an all-lab share, so it is held at the first quarter of 2025 — the last quarter before the dominant lab's results fall out of the captured record at Illinois's mid-2025 reporting transition (the dashed seam; see the note below). After that point the denominator is incomplete, and a share read off it would overstate craft. The lab-stable craft trend to the right carries the period past the seam.
The craft ceiling
January 2024: Illinois raised the craft canopy cap, 5,000 → 14,000 sq ft. Applications opened that March.
Built from the two labs whose reporting never broke at the mid-2025 data transition (ACT and Steep Hill), so this craft line runs clean and at full strength through the end of 2025 — see the note on the recent data in the Labs section. No craft grower tested flower until 2023. Active craft growers reached twenty by late 2025 — and still climbing. The volume sits low against the incumbent tier the whole way: lab-stable craft output ran from a handful of packages at the start to a few hundred a quarter — 540 in early 2025, 664 in the third quarter — with no break from the ceiling. More growers, same structure.
Each new cohort of growers, against the share it captured
Growers grouped by the year they first entered the testing record. Bars show how many growers each cohort added; the points show the share of all tested flower that cohort holds today. The 21 incumbent sites that entered in 2021 — run by 18 operators, since Cresco runs three and Green Thumb runs two — still hold 95.32% of the record. Every cohort since has added growers and captured almost none of it — not because they tested less per grower, but because the structure decided the ceiling before they entered.
Read across, not up: this is not a timeline. It is the same question asked of each entry class — how many came in, and how much of the market they were able to reach. The newest entrants (21 growers since 2023) together hold 4.7% of tested flower. Shares here are computed over the full record, which is overwhelmingly the 2021–early-2025 period; the mid-2025 record seam (see the note below) sits at the tail and does not move the incumbent tier off its 95.32%.
Across five years of the state's testing record, the craft tier sits at a fraction of the incumbent tier and never climbs out of it — 88.1% of tested flower from twenty-one cultivation-center sites run by eighteen operators, 11.8% from craft, with lab-stable craft volume reaching only a few hundred packages a quarter at its 2025 peak and no break from the ceiling. That is the structure, on the public record. The craft growers' own association explains why the structure is so hard to escape. The cost of breaking ground doesn't scale with the canopy you're cleared to grow. As Scott Redman, who runs the Illinois Independent Craft Growers Association, put it to the legislature: "The amount of infrastructure you have to build — the offices, the transport bay, the vaults — is the same whether you have 5,000 or 14,000 square feet to grow," and "all of that costs millions of dollars." (Illinois Answers, October 2023.)
In January 2024, Illinois pulled the only lever it had: it raised the craft canopy cap from 5,000 to 14,000 square feet, and opened applications that March. Do not read the flat line after that date as the expansion failing. You cannot read it at all yet — and that is the point. Adding canopy is not a setting you toggle. It is new construction, new capital, a new grow cycle, a harvest, and then the lag before a single extra package reaches a lab and surfaces in this data. The rule is just over two years old; applications opened twenty-seven months ago. By the physics of building a grow, nothing the expansion does — in either direction — can have shown up here yet. And that latency is the finding. The reason the gap persists isn't mystery; it's capital, named by the people closest to it. State Rep. La Shawn Ford, who chairs the Illinois House's cannabis working group, put it plainly as the cap fight came to a head: craft license holders, he said, were "on the brink of losing their licenses because they don't have the money to scale up." (NPR Illinois, March 2023.) A bigger canopy on paper does not write that check. The next few years are the literal tell, not a rhetorical one. The number worth tracking is whether lab-stable craft volume moves past its current few-hundred-a-quarter ceiling once the expanded canopy has had time to produce. That is what will show whether the expansion changed the structure or just the paper ceiling.
That is the data. Here is the reading. The incumbency in Illinois is insurmountable — by competition. That is the entire point. There is very little a market can do against a 15-to-1 statutory canopy advantage that holds even at full expansion. The only lever that moves a structure like this is the same one that built it: the state. Illinois did move it — and the people inside the tier will tell you why the move takes so long to land. Ambrose Jackson, who runs the craft grower Helios Labs and got his operation off the ground, describes the wall in dollars: "We had to raise half-a-million dollars up front to build out this space before we could deliver even one dollar of product." (Illinois Answers, October 2023.) That is the difference between a structure the state can still repair and one the market was ever going to fix on its own. The growers who built into the cap, and the legislators who tried to raise it, have been saying for years that the room is too small. That's the interpretation. The chart above, and the clock, are the facts.
A note on "unconcentrated"
Run the standard concentration test on Illinois and it comes back clean: an HHI of 831, well under the Justice Department's 1,500 threshold for an unconcentrated market. Ignore it. That number is answering a question Illinois never allowed to be asked. Concentration math assumes firms are free to enter and grow into whatever share they can win. Illinois caps the license, caps the canopy, and caps it again at expansion. When the state decides in advance how large each tier is permitted to become, a measure of how large each tier became is measuring the wrong thing. The HHI doesn't see the structure. It sees the result the structure was built to produce and calls it competition.
Year over Year
The pass rate climbed from 88.35% in 2021 to 98.8% by 2026 — a 10-point jump — while median flower THCA drifted from 24.4% to 28.2%. That is the shape of a market growing up: operators learn the panels, dial in the rooms, and stop sending out flower that fails. The named-lab data further down lets you see how the curve looks lab by lab.
Median flower THCA by year
Annual median for raw flower (Buds + ShakeTrim) only.
Pass rate by year
Share of completed packages that passed all required panels.
Failure rate by year
Package-level failure rate per calendar year. The mirror image of the pass-rate climb.
Volume tested by year
Completed package tests per year. 2021 and 2026 are partial years. 2025 reads low: the dominant lab's results fall out of the captured record at the mid-2025 reporting transition, so the second half of the year is undercounted — see the note in the Labs section.
Testing & Safety
95.78% of completed packages pass. 4,243 failed. When Illinois flower fails, it fails on one thing far more than any other: microbial contamination — yeast and mold accounts for 3,967 of the failing packages, against 20 pesticide failures and 1 heavy-metal. Potency is not a pass/fail criterion.
One scope note on that 95.78%. Two of the largest incumbent sites — Cresco's Lincoln facility and Green Thumb's Oglesby facility — and Deibel, the lab that tested Cresco, record zero or near-zero failures across very high volume. That is almost certainly a completeness gap in the public-records export, not a flawless production run. Their packages are counted in the volume and potency figures; their pass rates are treated as unverified and excluded from any per-operator finding. Read the true cultivation-center pass rate as modestly lower than the headline.
What actually fails
Packages with at least one failed result, by test category. Illinois records do not name the failing analyte, only the panel.
Pass rate by product type
Completed packages by product category.
| Category | Packages | Pass rate |
|---|---|---|
| Buds | 91,212 | 95.56% |
| ShakeTrim | 9,435 | 97.97% |
Flower (Buds and ShakeTrim) is essentially the entire dataset. Extracts and edibles appear in trivially small numbers and are not reported here.
Pass rate by license class, year over year
Illinois is the only state where every package can be tagged to its cultivation-center or craft-grower license by name. Once the craft tier shows up in volume (2023 on), the two tiers pass within roughly a point of each other every year. The license class governs volume; it does not separate the two on safety.
Craft's 2023 line (87.02%) reflects a single early grower's first quarters in the record, not the tier; by 2024 it tracks the incumbents within a point. Tiers carry different volume — read the rate, not the gap. The incumbent 2025 point sits on a partial year: the dominant lab's results fall out of the captured record at the mid-2025 transition, so the second half of the incumbent tier's volume is missing from that year's rate. Read 2025 with that caveat; the 2021–2024 trend is the durable signal.
Potency
THCA is the primary cannabinoid in raw flower. Median Illinois flower lands at 25.2%, drifting up from 24.4% in 2021 to 28.2% in 2026. The per-lab spread is real but modest — and part of it reflects which products each lab receives, not the lab itself.
THCA distribution
Flower packages by reported THCA %, 2% bins. 91,209 packages. Median 25.2%, mean 25.2%.
Where each lab falls on the THCA distribution
Density curves per lab (100+ flower samples) overlaid on the market distribution. Hover for the density at any THCA value.
Incumbent vs craft, on the same curve
License-class density overlay — an IL-only cut. The two tiers land within about a point of each other on median flower THCA (craft 27%, cultivation center 25.1%). The structure caps craft on volume — the potency curves sit on top of one another.
Median THCA by lab
Median reported flower THCA % at each lab. Labels show sample size.
Labs
Six labs are named in the record, but two — LK Pure and ACT — handle nearly 90% of all volume. Pass rates vary by several points, but labs do not test the same population of operators, so the spread requires context.
Lab market share
Share of completed packages, by testing lab.
Pass rate by lab
Share of packages passing, by lab. Low-volume labs flagged below.
Lab market share over time — where the record changes hands
Packages tested per lab per quarter, full quarters only. For most of the record two labs carry the market: LK Pure — which tested the incumbent tier almost exclusively — and ACT. Read the right edge. At Illinois's mid-2025 BioTrack→Metrc reporting transition, LK Pure's results fall out of the captured data, and ACT plus Steep Hill carry what remains. This is the seam itself, drawn. It is a change in what the record captures, not a finding about any lab's operations — and it is why the all-lab and incumbent series above are held at the first quarter of 2025.
A note on the recent data — the record caught its own seam
In mid-2025, Illinois moved its cannabis tracking from BioTrack to Metrc. At that transition, the results from LK Pure — the lab that handled the incumbent tier almost exclusively, roughly 60% of all testing — fall out of the captured record. You can see the moment in the chart above: the LK Pure band runs the length of the data and then drops to almost nothing at the seam. We cannot tell, from inside this dataset, how much of that is a reporting-pipeline change and how much is real. So we do not guess.
We do two things instead. The craft tier barely used LK Pure — it tested at ACT and Steep Hill, the two labs whose reporting carried through the transition intact. That makes the craft signal clean, and we show it at full strength, all the way through the end of 2025; that is the craft-ceiling chart at the top of the page. The all-lab and incumbent-tier series do depend on the volume that fell out, so we hold those at the first quarter of 2025 — the last quarter before the seam — and mark the cut directly on the chart rather than carry a denominator we know is incomplete. A share read off the post-seam record would overstate craft, because the gap sits almost entirely on the incumbent side. We will not show you that number.
This is the same posture as the other flags on this page. Where two incumbent sites and one lab record implausibly clean results, we count their volume but hold their pass rates as unverified. Where the record loses a lab at a system transition, we show what survived intact and hold what didn't. The honest move is to mark the seam, not paper over it — so that is what the charts do.
Lab profiles
Click a tab for each lab's volume, pass rate, median potency, and where it sits against the market median.
All labs
Full named-lab table. Sample size matters: read the small labs with care.
| Lab | Packages | Share | Pass rate | Median THCA | Active |
|---|---|---|---|---|---|
| LK Pure Labs | 60,472 | 60.08% | 96.24% | 24.67% | Sep 2021 – Dec 2025 |
| ACT Laboratories, Inc. | 30,032 | 29.84% | 94.53% | 26.67% | Sep 2021 – Feb 2026 |
| Smithers CTS IL, LLC (Previously Origo Labs, LLC) | 4,822 | 4.79% | 94.59% | 23.57% | Sep 2021 – May 2025 |
| Steep Hill Illinois | 4,163 | 4.14% | 99.02% | 25.94% | Dec 2023 – Feb 2026 |
| Deibel Bioscience of Illinois Inc | 1,132 | 1.12% | 100% | 22.55% | Sep 2021 – May 2022 |
| Grace Analytical Laboratory, Inc ⚠ | 26 | 0.03% | 15.38% | 20.65% | Feb 2022 – Dec 2023 |
⚠ Grace Analytical (26 packages) and Deibel Bioscience (a single-client lab, 2021–2022) carry too little volume to draw lab-wide conclusions. Their pass rates reflect one operator's results, not the lab's standards.
Failure Analysis
4.22% of completed packages fail — and that rate has collapsed over the program's life, from 11.65% in 2021 to 1.2% in 2026. But the program-wide number hides an enormous spread between operators: from 26.61% at the top to 10 operators with a perfect record. Illinois names every operator, so this is attributable — not anonymized.
Overall pass / fail
100,647 completed packages. Package-level overall result.
Failure rate by year
Package-level failure rate per calendar year — the clearest downtrend in the dataset.
Failure rate by operator — the spread the headline hides
Every named operator, ranked by package failure rate. The program average is 4.22%, but operators range from 26.61% down to a hard zero. Same market, same labs, same rules — wildly different outcomes. Bars over the program average are drawn in rust.
⚠ The cluster of operators reporting 0% across very large volumes (Cresco Lincoln, GTI Oglesby) is implausible as a true result and almost certainly reflects a completeness gap in the public-records export — the failing packages simply weren't in the file we received — rather than a flawless record. The same gap shows up at Deibel, the lab that tested Cresco. This is a limitation of the data export, not a claim that any company concealed anything. Flagged, not dropped, and held out of every per-operator pass-rate finding. Note this is overall pass/fail — Illinois records do not carry the failing analyte, so a contaminant-level breakdown is not possible.
Failures don't track license class
The two-tier split that governs volume does not govern safety. Cultivation centers fail 4.27% of packages (4,099 of 95,934); craft growers fail 3.06% (144 of 4,711). The worst operators in the variance chart above sit inside the incumbent tier, not the craft tier. The structure caps how much craft can grow — it has nothing to say about how clean anyone's flower is. That is a per-operator story, and the public record names every one of them.
A note on the two craft totals on this page: the safety figures here count every craft package on record (4,711, all-data). The market-share section earlier counts only full quarters (4,401), dropping the partial periods at the edges of the data so each point on those charts is a complete three months. Same metric, two bases — not a contradiction.
Operators
Every named operator in the record, tagged to its license class. Search, sort, and open any one to see when it entered, how much it has tested, where it sends its product, and how it performs.
Tap an operator above to open its detail.
Download the Data
The whole point of this is that it's the public record, made usable. Here's the cleaned, deduplicated, package-level dataset behind every chart on this page — 100,647 rows. Check our math.
Package-level dataset (CSV)
One row per completed package: operator, lab, date, category, pass/fail, remediation flag, THCA.
Download il_packages.csv →How this was built
Source files, the license-class tagging, dedup rules, the finished-test filter, and every known caveat.
Read the methodology →Data through February 2026 · built June 2026.