"Non-Detect" Does Not Mean Zero

When a certificate of analysis reports ND, it means the analyte was not detected by that method at that sensitivity — not that its concentration is zero. Every analytical method has a floor, and ND is a statement about the floor as much as about the sample.

This is the most consequential piece of small print in the whole category, because “THC: ND” is routinely read as “contains no THC” by people whose jobs depend on the difference.

The two limits that define what ND means

Reports use two related figures, usually in a footnote or a column of their own.

LOD — limit of detection. The lowest concentration the method can reliably distinguish from background noise. Below this, the instrument cannot tell the difference between a small amount and nothing.
LOQ — limit of quantitation. The lowest concentration at which the laboratory is willing to state a number. Between the LOD and the LOQ, the analyte is detectable but not reliably measurable.

That gives three possible outcomes for any row, and reports distinguish them:

ND — below the LOD. Not seen.
< LOQ, sometimes written as “trace” or “detected, not quantified” — above the LOD, below the LOQ. Present, amount unreliable.
A number — at or above the LOQ. Present, quantified.

An ND and a < LOQ are different results, and a report that collapses them into one symbol has given you less information than one that does not.

Why a low limit is a stronger claim than a low result

Consider two reports, both showing ND on the THC row. One states an LOD near the bottom of what current instrumentation achieves. The other states an LOD an order of magnitude higher. The second report’s ND is compatible with a THC concentration that the first report would have flagged.

Same two letters. Very different statements about the batch.

So the useful reading habit is: never read a result column without reading the limit column. An ND with no stated limit is an assertion; an ND with a stated limit is a measurement with a known boundary. This is why cropped screenshots and low-resolution images of certificates are a problem in themselves — the footnotes are usually the first thing lost, and the footnotes are where the limits live.

Where labs get their limits from

Detection limits are properties of the method, the instrument, and the sample matrix — not arbitrary choices. A cannabinoid panel run by liquid chromatography on a clear oil will have different limits from the same panel run on a gummy, a cream, or a plant sample, because the surrounding material interferes differently.

Two practical consequences:

A limit that applied to one product format does not carry over to another. A report on an extract is not a report on the finished bottle, which is why the sample description field matters — see how to read a certificate of analysis.
Limits vary between accredited laboratories doing competent work. Different limits are not evidence of a bad lab. Undisclosed limits are a different matter.

The rows where this bites hardest

THC. Obviously. A “THC-free” or “non-detectable THC” claim rests entirely on a detection limit, which is the argument made at length in what THC-free can and cannot mean. Add the complication that a report may carry separate delta-9, THCA, and total THC rows, each with its own result — so delta-9 ND does not imply total THC ND. That relationship is in total THC vs. delta-9 THC.

Pesticides. Panels typically report each compound against an action limit with a pass or fail. Most rows read ND on a clean sample. The meaningful question is how many compounds were screened and at what limits, not whether the column is full of NDs.

Heavy metals. Hemp takes up metals from soil, so this panel is particularly relevant for this crop. Here the limits are usually set against a regulatory specification, and ND against a stated specification is a stronger result than ND against nothing.

Residual solvents. Directly connected to the extraction claim on the front of the package — see extraction words on a label. A CO2 extraction claim is entirely compatible with a residual ethanol figure, because downstream refinement steps use ethanol.

Minor cannabinoids. On an isolate report, ND across the minor cannabinoid rows is the expected pattern and confirms the category. On a full-spectrum report, the same pattern contradicts the label. Same result, opposite meaning depending on the claim it is checked against — the point made in what CBD isolate means and in the spectrum categories post.

What ND does license you to conclude

Being fair to the documents: an ND from a competent laboratory at a disclosed low limit is genuinely useful information. It says the analyte, if present at all, is present below a known concentration. That is a real and checkable statement, and it is far more than a front-of-pack claim gives you.

What it does not license is the word “zero”, or the word “none”, or a conclusion about what will happen in a test or in a body. Whether a known upper bound is low enough for your circumstances is not a labelling question. If a testing programme or a medication is in the picture, that is a conversation for the programme administrator, an occupational health professional, or a pharmacist — several of the questions worth raising are listed in questions to ask a pharmacist.

Reading habits that make the difference

Read result and limit as a pair. Never one without the other.
Distinguish ND from < LOQ. They are different findings.
Check the units on the limit — µg/g, mg/g, % w/w. An unfamiliar unit is a common way to misjudge a limit by orders of magnitude.
Find the sample description and confirm the report is on the format you are holding, not on an upstream extract.
Insist on the full document. If the footnotes have been cropped away, the limits have been cropped away with them.
Match the batch code. A report for another batch describes another batch, however good its limits are.
Do not upgrade ND to zero in your own head when repeating what the report said. That upgrade is where most bad decisions in this category start.

Eased is a consumer-information site about product labelling. Nothing here is medical or legal advice, and nothing here describes what CBD does or does not do in the body. Talk to a doctor or pharmacist before using any CBD product, especially if you take other medication, and check the current law where you live.