Why Two Labs Can Disagree on Results

If you’ve ever seen different CBD lab results for the very same batch, you’re not alone. Small but noticeable differences can appear between Certificates of Analysis (COAs), and sometimes those gaps feel big enough to raise questions. Understanding why this happens helps you compare reports confidently without losing sight of what matters: product safety and consistency.

In THC-free products, even tiny differences in lab method or reporting can change how results look—especially for THC, where “non-detect” depends on the lab’s limit of quantitation (LOQ). The goal here is to explain common sources of variance and what you can do as a customer to interpret COAs sensibly.

This page focuses on why results differ. For a broader overview, see our Safety & Testing hub. For step-by-step reading guidance, see How to Read COAs (Lab Reports). For standards guidance, see Top Lab Testing Standards.

What “THC-free” and “non-detect” mean in a COA

“THC-free” products are typically formulated so that Delta-9 THC is not detected at or above a lab’s LOQ. Non-detect (ND) does not mean absolute zero—it means the analyte was below the lab’s validated ability to measure it precisely. Two labs can have different LOD/LOQ values. One lab may report ND while another shows a small “<LOQ” number. Both can be technically correct.

Key points:

ND = below the lab’s reporting threshold, not necessarily zero.
Different LOQs across labs can create apparent discrepancies for THC, CBD, and minor cannabinoids.
COAs may report Delta-9 THC only, “total THC,” or both. Check which is used.

The main reasons two labs may disagree

Method and instrument differences

Chromatography choices: HPLC-UV/DAD vs LC-MS can differ in selectivity and sensitivity.
Gas chromatography may convert cannabinoid acids to neutrals unless derivatized, affecting totals.
Validation scope: labs validate different matrices and ranges, which can shift bias and uncertainty.

Sample prep and matrix effects

Different extraction solvents, dilution factors, and filtration can change recoveries.
Matrices (oils, gummies, beverages, topicals) interact differently with methods; emulsifiers, sugars, waxes, and essential oils can suppress or enhance signals.
For topical-specific considerations, see Essential Oil Safety in Topicals.

Sampling and homogeneity

Where the sample came from matters: top, middle, or bottom of a tank; one gummy vs a composite of many.
Grinding and homogenizing solid samples reduces sub-sample error; skipping it can skew results.
Moisture differences (especially in edibles) change mg/g potency calculations.

Reference standards and calibration

Purity and source of reference standards can differ across labs.
Calibration range, curve fitting, and internal standards impact accuracy at low levels (important for THC-free claims).

LOQ/LOD and reporting rules

Some labs round “0.004%” to ND; others report “<LOQ.”
Panels vary: a lab may include more minor cannabinoids than another, changing totals and ratios.
Units differ: mg/g, mg/mL, mg/serving, or % by weight. Always align units before comparing.

Storage, age, and stability

Heat, light, oxygen, and time can change cannabinoid profiles (e.g., THC can oxidize to CBN).
If samples were collected weeks apart, natural drift may explain variance. See Stability Testing.

Accreditation and quality systems

ISO/IEC 17025 accreditation indicates a validated quality system and method competence but does not guarantee identical results across labs.
Measurement uncertainty (MU) quantifies the expected spread around a result; two labs can be different yet both fall within MU. See Top Lab Testing Standards.

How to compare COAs wisely

Match the batch/lot number, product name, flavor/strength, and matrix.
Check the sample collection and report dates; older vs newer samples may differ.
Align units and serving sizes; convert if necessary (mg/g vs mg/mL vs mg/serving).
Confirm whether the lab reports “Delta-9 THC,” “total THC,” or both.
Look for the LOQ and how “ND” is defined; note any “<LOQ” annotations.
Scan for method notes and measurement uncertainty; small differences may be within MU.

When a difference is a red flag

Order-of-magnitude (10x) differences without a clear explanation.
Mismatched lot numbers or product descriptions.
No LOQ/LOD, no method description, or no analyst signature/approval.
Refusal to share basic method or LOQ info on request.

THC-free products and drug testing

Employer drug tests vary in technology and cutoffs. A product labeled THC-free generally targets non-detect Delta-9 THC at the lab’s LOQ, which is not the same as a guarantee for every drug test. Different tests and sensitivities exist, and cross-reactivity is possible. If testing is a concern, review the COA carefully and consider discussing with your employer or testing provider.

What to ask a brand or lab

Was the sample a composite from multiple units or a single unit?
What are the LOQ values for Delta-9 THC and “total THC” in this matrix?
Which analytical method and instrument were used?
Is the lab ISO/IEC 17025 accredited for this method and matrix?
What is the stated measurement uncertainty?
Can I see the most recent COA for my exact lot?

Disclaimer

This content is for general information only and is not medical, legal, or testing advice. Product compositions and lab capabilities vary. Always review the COA for your specific lot and consult qualified professionals if you have questions about regulations, drug testing, or personal health.

FAQ

Why do two labs give different CBD results?
Differences in sampling, methods, instruments, LOQs, and reporting rules commonly create small to moderate variance.

Which lab is “right”?
Both can be valid within their measurement uncertainty. Ask for LOQs, methods, and uncertainty to contextualize results.

What is an acceptable variance?
It depends on matrix, method, and MU. Small percentage differences are common; large discrepancies warrant follow-up.

Can one lab show THC as ND while another shows a trace?
Yes. Different LOQs or reporting rules can produce ND in one report and “<LOQ” in another.

How do I resolve COA discrepancies?
Confirm batch match, align units, request LOQs and methods, and consider a third-party retest with a composite sample.

Does ND guarantee I’ll pass a drug test?
No. ND reflects the lab’s LOQ, not all possible test cutoffs or methods.

Summary

Different CBD lab results usually stem from normal scientific factors: sampling, method choices, instrument sensitivity, LOQs, and reporting conventions. Aligning units, verifying batch details, and understanding ND/LOQ will explain most COA discrepancies. When in doubt, ask about methods, LOQs, and measurement uncertainty—or seek a confirmatory retest.

Looking for options formulated and tested with THC-free goals in mind? Explore products at https://www.cannagea.com/thc-free.