Most operators experience the cannabis testing process as a black box. A sample goes in, a report comes back some days later, and what happened in between is unclear. That gap is where a lot of unnecessary friction lives — failed batches nobody can explain, disputes about turnaround, and results that arrive without context.
Here is what actually happens, in order. If you want the wider context first, start with what cannabis lab testing is and the breakdown of what each analysis measures.
Step 1: Sampling
Everything downstream depends on this step, and it is the one most likely to be underestimated.
A batch is defined and declared: a specific quantity of material, from a specific harvest or production run, with a batch identifier. A sampler then pulls increments from multiple points across that batch according to a defined protocol — not a handful from the top of one container.
The reason is straightforward. Cannabis is heterogeneous. Cannabinoid concentration varies by plant position, and microbial contamination is often localised. A sample drawn from a single point describes that point. A composite sample drawn systematically across the lot describes the lot.
In most regulated markets, sampling is performed by the laboratory or by a licensed third party rather than by the producer, specifically to remove the incentive to select favourable material.
Step 2: Chain of Custody
From the moment the sample is taken, it is documented. Who collected it, when, from which batch, in what quantity, under what storage conditions, and every transfer of possession after that.
Chain of custody is not bureaucracy for its own sake. It is what makes a result defensible. If a report is ever questioned — by a regulator, a buyer, or in a dispute — the custody record is the evidence that the material analysed is the material declared. A result without an intact custody trail is difficult to stand behind, regardless of how good the analysis was.
Step 3: Intake and Accessioning
At the laboratory, the sample is logged into the information management system and assigned an internal identifier. Condition on arrival is recorded — temperature, packaging integrity, whether the quantity matches the paperwork.
Samples can be rejected at this stage. Insufficient quantity for the requested panel, damaged packaging, a mismatch between the declared and delivered material, or missing documentation are all legitimate grounds. It is better to reject at intake than to produce a result nobody can rely on.
The requested panel is confirmed here too: which analyses apply to this product type, and whether the jurisdiction requires anything beyond what was ordered.
Step 4: Sample Preparation
Raw cannabis cannot be fed directly into an instrument. Preparation converts it into a form the analysis can work with, and it varies by test.
For potency, material is homogenised and a precise mass is extracted into solvent, then diluted to a concentration within the instrument’s calibrated range. For heavy metals, the sample is digested in acid to break down organic matter and release the elements for measurement. For microbial testing, the sample is prepared under aseptic conditions to avoid introducing organisms during handling.
Preparation is where a significant share of analytical error originates. Incomplete homogenisation, extraction that does not fully recover the target compounds, or contamination introduced in the lab all produce results that look fine on the report and are wrong.
Step 5: Analysis
Different questions require different instruments.
- Cannabinoids are typically measured by high-performance liquid chromatography (HPLC), which separates compounds in solution without heat — important, because heating would convert THCA to THC mid-analysis and distort the result.
- Terpenes are volatile, so gas chromatography is the usual approach, generally with mass spectrometry for identification.
- Residual solvents use headspace gas chromatography, sampling the vapour above the material rather than the material itself.
- Pesticides require tandem mass spectrometry to detect trace residues across a long list of compounds at very low concentrations.
- Heavy metals are measured by inductively coupled plasma mass spectrometry (ICP-MS), which can quantify elements at parts-per-billion levels.
- Microbials use either culture-based methods or quantitative PCR, depending on the organism and the regulatory requirement.
Each run includes quality controls: calibration standards, method blanks to confirm nothing is being introduced by the process, and spiked samples to verify the method is recovering what it should. If the controls fail, the batch of analyses is repeated. More on the instrumentation side in how technology is changing cannabis lab testing.
Step 6: Data Review
Instrument output is not a result. An analyst reviews the chromatography, checks that peaks are properly integrated and resolved, verifies that quality controls passed, and confirms the calculations that convert raw signal into a reported concentration.
In a well-run laboratory this review is performed by someone other than the analyst who ran the sample. Second-person review catches integration errors, transcription mistakes and results that are technically valid but implausible for the material.
Anything unusual gets investigated before it gets reported — an unexpected peak, a value far outside the range typical for the product type, a control that drifted. Reporting first and investigating later is how bad data gets into the market.
Step 7: Reporting
The certificate of analysis is generated: batch identifier, sample details, methods used, results per analyte, applicable action limits, pass or fail status, and the date of analysis. Limits of detection and quantitation should be stated so that a “not detected” result can be interpreted properly.
The COA is signed by an authorised person and released. Once released it is a formal record — corrections require a documented amendment, not a quiet edit. Our guide to reading lab results walks through each section of the finished report.
Step 8: Retention
Sample retains, raw instrument data and records are kept for a defined period. If a result is challenged, retained material can sometimes be re-analysed, and the underlying data can be reviewed independently.
Retention is also what makes a recall workable. When a problem surfaces after release, batch-level records are what identify the scope.
How Long It Takes, and Why
Turnaround varies by panel. Potency and terpenes are comparatively quick. Culture-based microbial methods take days because organisms need time to grow — that is biology, not backlog. Pesticide panels are among the slower analyses because of preparation and run time across a long compound list.
A laboratory quoting a turnaround dramatically shorter than the industry norm for a full panel is worth a question. Some steps have a floor, and the ways of getting under it involve skipping something.
What Producers Can Do to Get Cleaner Results
Several common problems are avoidable on the production side rather than in the laboratory.
- Declare batches accurately. An overstated batch size makes the sampling plan unrepresentative before anything else happens.
- Submit enough material for the full panel, including any retest. Short samples cause delays and sometimes rejections.
- Verify your cure before submitting. Moisture and water activity problems are the most common avoidable failures, and more drying time is cheaper than a failed batch.
- Keep records of every input that touched the crop. When a pesticide result comes back positive, that record is how you find the source.
- Store samples appropriately between collection and delivery. Heat and humidity change material.
A laboratory that will walk you through where a problem originated is worth considerably more than one that emails a result and stops — one of the criteria in how to choose a cannabis testing laboratory.
The process is designed so that each step constrains the next. Good sampling makes analysis meaningful, good preparation makes instrumentation meaningful, and good review makes the report meaningful. Skip any one of them and the certificate is a document rather than a measurement.