How Often Should Kratom Be Lab Tested? Batch, Lot and Risk-Based Testing Explained

Short answer

There is no single scientifically responsible answer such as “test
kratom every 30 days” or “test it once per year.” The useful testing
frequency depends on what is being tested, what material is being
represented, how the lot was made, what could have changed and which
specification or legal threshold is being evaluated.

A defensible program may include:

• identity work on each unique incoming botanical lot;
• risk-based contaminant testing on incoming material or finished
  lots;
• in-process checks at meaningful manufacturing steps;
• finished-batch verification under a written sampling plan;
• packaging and label checks for each packaged lot;
• periodic stability studies that support a best-by or expiration
  date; and
• additional testing after a supplier, process, laboratory, formula,
  storage condition or regulatory requirement changes.

Those are different activities. A seller who says only “lab tested”
or “batch tested” has not told you which activity occurred.

The most useful consumer question is therefore not simply, “Was this
tested?” It is:

Which lot was sampled, who collected the sample, what panel
was run, when was it run, and does the report match the product being
offered now?

Why testing frequency
is not one number

Testing is a way to answer a defined question. Different questions
require different samples, methods and timing.

For example:

• Identity: Is this material actually consistent with
  Mitragyna speciosa rather than a different plant or
  substitute?
• Composition: What concentrations of mitragynine,
  7-OH or other reported alkaloids were measured?
• Microbiology: Did the tested sample meet the
  written limits for specified organisms or microbial counts?
• Elemental contaminants: What did the sample contain
  for lead, arsenic, cadmium and mercury?
• Pesticides: Which compounds were included in the
  laboratory panel, and what were the reporting limits?
• Residual solvents: If extraction solvents were
  used, what residues were measured in the finished extract?
• Uniformity: Does a blend, capsule run or packaged
  lot remain reasonably consistent across locations or units?
• Stability: Does the product continue to meet
  specifications through its labeled shelf life?
• Legal threshold: Does the result, unit and
  denominator allow comparison with the rule that applies in the
  destination jurisdiction?

An identity result does not answer a lead question. A four-metal
panel does not answer a Salmonella question. A potency result
does not prove the sample was representative. A report generated twelve
months ago does not automatically describe a newly received lot.

Testing frequency should follow the question and the risk—not a
slogan.

First define “batch,”
“lot,” “sample” and “test”

Testing programs become confusing when basic terms are used
loosely.

Batch

A batch is a specific quantity produced during a defined
manufacturing cycle and intended to have reasonably uniform character
and quality. A batch might be one blender load, one encapsulation run or
one extraction campaign. The exact boundary should be written down
before sampling.

Lot

A lot is a quantity identified for tracking and control. A received
botanical lot may come from one supplier shipment. A finished-product
lot may combine material from more than one incoming lot or divide one
manufactured batch across several packaging runs. “Lot” and “batch” are
sometimes used interchangeably in casual speech, but a quality record
should state what each code represents.

Sample

A sample is the portion taken for examination or testing. The
laboratory usually sees only a small fraction of the total lot. How that
fraction was selected, combined, sealed, transported and received can
matter as much as the instrument used to analyze it.

Test

A test is a defined examination or analytical procedure. “Tested”
might mean a visual examination, microscopy, an alkaloid assay, a
microbial panel or a metals panel. It does not necessarily mean all of
them.

Test panel

A panel is a group of analytes or examinations. Even panel names can
hide major differences. One “pesticide panel” might cover dozens of
compounds while another covers hundreds. One “microbial panel” might
include total counts but omit specified pathogens. The analyte list,
method and reporting limits matter.

What federal
quality rules teach about frequency

FDA’s current position is that kratom is not lawfully marketed as a
dietary supplement. For that reason, Kiody should not describe
compliance with dietary-supplement current good manufacturing practice
as FDA approval, lawful-market status or a federal kratom
certification.

Even so, the structure of 21 CFR Part 111 is a useful quality-system
reference because it separates several types of verification instead of
prescribing one calendar interval for every test.

FDA’s Small Entity Compliance Guide explains that manufacturers
operating under Part 111 must establish whether specifications are met
for components, in-process production, finished batches, received
products, packaging and labels. It also explains several distinct
approaches:

• dietary-ingredient identity testing is generally conducted on each
  unique lot received, unless FDA grants a qualifying petition for an
  alternative;
• other component specifications can, under defined conditions, rely
  on a qualified supplier’s certificate of analysis, with initial
  confirmation and periodic reconfirmation of supplier reliability;
• in-process samples are collected from each manufactured batch at
  points where control is necessary;
• finished-batch verification may cover every batch or a statistically
  selected subset under a sound sampling plan, but selected specifications
  must provide adequate assurance that the complete product specifications
  are met;
• packaged and labeled lots require appropriate examination; and
• quality-control personnel review laboratory operations and
  results.

This is not a simple “test everything on every batch” rule, nor is it
permission to avoid meaningful verification. It is a system of written
specifications, scientifically valid methods, representative sampling,
documented supplier qualification, quality review and justified
decisions.

For consumers, the practical lesson is simple: a credible testing
program describes its logic. Frequency should be documented, connected
to a specification and revised when evidence changes.

The seven layers of a
testing program

1. Incoming botanical
identity

Identity testing asks whether an incoming botanical material is what
the supplier says it is. This question usually belongs at receipt,
before material is released for use.

Identity may involve macroscopic or microscopic examination, chemical
fingerprints, DNA methods, chromatographic profiles or a scientifically
justified combination. The right method depends on the form. DNA can
become less informative after extraction or extensive processing.
Microscopy may be useful for intact or powdered botanical material but
cannot by itself quantify contaminants or alkaloids. An alkaloid result
can support a chemical profile without proving every aspect of botanical
identity.

Changing the supplier, country or region of origin, plant-part
description, harvest practice, milling process or visual character
should prompt closer review. A result for one incoming lot should not
automatically be carried forward to a later shipment simply because the
product name is unchanged.

2. Incoming
contaminant and composition testing

Some hazards may enter with the botanical material. Depending on the
risk assessment, incoming-lot testing may include microbiology,
elemental contaminants, pesticides, mycotoxins, moisture or water
activity, foreign material and alkaloid composition.

The panel should reflect the material and supply chain. Agricultural
origin, drying conditions, known regional pesticide use, prior supplier
performance, shipping conditions and historical results can all affect
the plan. A supplier certificate can be useful evidence, but it should
not be treated as self-proving. The buyer should verify the lot
connection, laboratory, method, units, limits and authenticity.

3. In-process checks

Manufacturing can introduce or change risks even when incoming
material passed testing. Examples include:

• blend segregation;
• cross-contact from shared equipment;
• incorrect component addition;
• capsule-fill variation;
• exposure to moisture;
• contamination after a microbial-reduction step;
• extraction or solvent-removal problems; and
• packaging-line mix-ups.

Some in-process controls are measurements rather than external
laboratory panels. Blend time, sieve integrity, capsule weights,
metal-detector challenges, line clearance and seal checks may be more
useful at the point where the process could fail. Waiting for a finished
COA cannot reconstruct every missed process control.

4. Finished-batch
verification

Finished-batch testing asks whether the manufactured product meets
selected specifications before release. It may confirm alkaloid
composition, microbiology, contaminants, moisture or other attributes.
The selected tests should be connected to all relevant finished-product
specifications, either directly or through a documented quality
rationale.

A program may test every finished batch for certain critical
attributes. In other circumstances, it may use a statistically justified
subset of batches and rely on validated process controls plus other
verification. The phrase “periodic testing” is not a complete
justification by itself. The business should be able to explain the
population being sampled, selection method, frequency, historical
evidence and escalation triggers.

5. Packaging and label
verification

Testing does not stop when bulk material passes. Each packaging run
can create new errors: the wrong label, an unreadable lot code, an
incorrect net quantity, a damaged seal or a mismatched COA link.

Appropriate examinations may include label reconciliation, barcode
and lot-code checks, net-content controls, seal-integrity checks and
confirmation that the public COA points to the correct finished lot.
These controls may occur on each packaged batch or at defined intervals
during a run. Their frequency should be sufficient to detect a change
before a large quantity is affected.

6. Stability and shelf-life
work

A release COA describes a sample at a point in time. It does not
establish how long the product remains within specification.

Stability studies examine change over time under defined packaging
and storage conditions. Real-time studies are the strongest direct
evidence for the labeled period. Accelerated studies can support
development or risk assessment but do not automatically substitute for
real-time data. Opened-package or in-use studies may be important when
repeated exposure to air, light or moisture is foreseeable.

The stability schedule is not the same as routine batch testing. It
uses planned time points and representative lots to support a shelf-life
conclusion. A new formula, packaging system, storage condition or
extract concentration can require a new or bridged stability
assessment.

7. Triggered and
investigative testing

Extra testing should occur when evidence creates a new question.
Triggers may include:

• a complaint or adverse-event report;
• a failed, atypical or near-limit result;
• damaged or wet shipments;
• loss of temperature or humidity control;
• a supplier, farm, processor or laboratory change;
• a new extraction solvent or manufacturing step;
• a sanitation failure or environmental finding;
• unexpected color, odor, texture or capsule-weight variation;
• a regulatory alert, recall or changed legal threshold;
• an unexplained trend across recent lots; or
• a suspected label or lot-code mix-up.

Triggered testing should be designed around the problem. Repeating
the same small panel may not answer the new question.

A practical frequency matrix

The table below is an educational framework, not a universal
prescription. A qualified quality team should adapt it to the product,
process, history and applicable law.

Control question	Typical decision point	Frequency logic	What can increase frequency
Botanical identity	Incoming component release	Each unique incoming botanical lot under the Part 111 benchmark	New supplier, changed origin, atypical appearance, processing change
Supplier COA confirmation	Supplier qualification and ongoing monitoring	Initially confirm; periodically reconfirm under a written schedule	Discrepancies, failures, method changes, missed documents
Alkaloid composition	Incoming or finished-lot release	Lot- or batch-linked based on specifications and product type	Extracts, enhanced products, legal thresholds, near-limit results
Microbiology	Incoming or finished release	Risk-based by lot/batch and process controls	Wet shipment, sanitation issue, high counts, post-treatment exposure
Elemental contaminants	Incoming or finished release	Supply-chain and historical-risk based	New geography, soil/source change, trend upward, supplier change
Pesticides	Incoming material verification	Based on agricultural controls, origin and panel relevance	New farm/region, new pesticide information, incomplete supplier panel
Mycotoxins	Incoming or finished release	Based on moisture, drying/storage history and supplier performance	Moisture excursion, mold concern, new harvest or storage issue
Residual solvents	Finished extract release	Tied to extraction process and solvent use	New solvent, process change, recovery issue, new extract supplier
Blend/capsule uniformity	During and after manufacture	At defined locations/times for each relevant batch	Formula change, segregation, fill-weight drift, equipment change
Packaging and labels	Packaging run release	Each packaged/label lot, with in-run checks as needed	Line changeover, new packaging, seal failures, wrong-label history
Stability	Planned study time points	Written protocol across representative lots	New package, formula, process, storage claim or adverse trend
Investigative testing	After a trigger	Whenever evidence raises a new material question	Complaint, OOS, regulatory action, recall, damage or excursion

The matrix deliberately avoids fixed numbers such as “quarterly” or
“annually.” Calendar intervals can be part of a plan, but the interval
needs a rationale.

How panel frequency
changes by product type

Whole leaf and plain powder

For ordinary botanical material, incoming identity and
agricultural-contaminant controls are central. Milling and blending can
change sample representativeness, introduce cross-contact or spread a
localized contaminant. Finished-lot verification should be connected to
the actual lot offered for sale.

Natural variation does not remove the need for specifications. It
makes representative sampling and trend review more important.

Pure-leaf capsules

Capsules add manufacturing questions even if the fill is only
botanical powder. The program should distinguish the weight of the empty
shell from the botanical fill, monitor fill consistency and control
label claims. A powder COA may support the capsule lot only when records
defensibly connect that powder to the encapsulation batch and show that
later processing did not create an unaddressed risk.

Extracts

Extraction changes the matrix and can concentrate both desired and
undesired constituents. It can also introduce solvent residues or
process-related variation. An extract program may therefore require
composition, residual-solvent and uniformity controls that would not
apply in the same way to plain leaf.

An extraction ratio such as 10:1 does not establish an alkaloid
concentration and cannot substitute for a quantitative result.

Enhanced products

“Enhanced” generally signals that an extract or isolated constituent
was added to botanical material. The manufacturer needs controls for the
added component, addition amount, blend uniformity and finished
composition. Testing only the base powder is not enough to characterize
the final enhanced product.

Concentrated 7-OH products

Concentrated or fortified 7-OH products are materially different from
ordinary botanical leaf. They require separate legal and analytical
review. Kiody does not sell concentrated 7-OH.

As of September 3, 2026, the federal proceeding concerning 7-OH above
a specified threshold remains pending, with the public-comment period
extended through September 10, 2026. That proposal is not a final
scheduling order. State and local restrictions may independently use
dry-weight percentages, percentages of total alkaloids, milligrams per
serving or categorical prohibitions. A result is useful for legal review
only if its units and denominator correspond to the applicable rule.

Manufactured derivatives

Mitragynine pseudoindoxyl (MGPI), MGM-15 and MGM-16 are temporarily
in federal Schedule I from August 26, 2026, through August 26, 2028,
unless the order is extended or permanent scheduling occurs sooner.

The Justice Department stated on September 1, 2026, that it would
exercise enforcement discretion when only incidental trace MGPI is
confirmed in an otherwise botanical-kratom product. That statement is
not a legal exemption, provides no numerical safe harbor and does not
apply to MGM-15, MGM-16, or manufactured, concentrated, fortified or
intentionally added MGPI. A laboratory program should not convert that
policy into an invented pass/fail number.

Does every lot need every
test?

Not necessarily—but every omitted or reduced test needs a defensible
control strategy.

The strongest reasons to reduce a particular test are based on
accumulated evidence: a qualified supplier, confirmed COA reliability,
consistent historical data, a controlled process, scientifically valid
methods, representative sampling, stable trends and written escalation
rules. Cost or convenience alone does not demonstrate control.

Some tests address hazards that are strongly lot-dependent. Some are
better placed at an earlier component stage. Some process attributes are
monitored continuously or several times during a run. Some destructive
tests are performed on a statistically selected subset. A well-designed
program can combine those controls without pretending that one finished
COA answers every question.

Conversely, “we have used the supplier for years” is not enough.
Relationships do not replace data. Supplier facilities, upstream farms,
laboratories, methods and processes can change without a familiar
company name changing.

When should testing
frequency increase?

A program should become more intensive when uncertainty or
consequence rises. Common reasons include:

1. A new supplier, broker, processor, farm or country of origin.
2. A supplier whose COA does not match independent confirmation.
3. A new laboratory, method, instrument platform or reporting
   limit.
4. A new extract, enhanced formula or manufacturing process.
5. A change in packaging, storage or shipping conditions.
6. A failed, invalid, atypical or near-specification result.
7. A complaint involving odor, appearance, foreign material or
   inconsistent units.
8. A microbial-reduction, sanitation or environmental-control
   failure.
9. A trend moving toward a limit even though individual results still
   pass.
10. A new state, local or federal threshold that requires different
    analytes, units or decision rules.
11. Incomplete chain-of-custody or uncertainty about what lot the sample
    represents.
12. A recall, enforcement notice or credible government alert affecting
    a supplier, region or product type.

Increased frequency can mean more lots tested, more increments within
a lot, additional analytes, lower reporting limits, confirmation by a
second laboratory or testing at a different process stage. Those are
different responses and should be selected deliberately.

When can frequency be
reduced?

Reducing frequency may be reasonable when a documented review shows
sustained control. A decision should consider:

• enough historical lots to evaluate variability;
• results from independent confirmation, not only supplier
  paperwork;
• supplier audit and change-notification performance;
• stable manufacturing and packaging processes;
• method suitability for the actual matrix;
• adequate reporting limits relative to specifications;
• absence of unresolved complaints, failures or adverse trends;
• a sound statistical sampling plan; and
• clear conditions that return the program to intensified
  testing.

A reduced schedule should have an owner, approval date and review
date. It should not continue indefinitely by inertia.

Identity deserves special caution. Under the Part 111 benchmark,
dietary-ingredient identity testing applies to each unique incoming lot
unless FDA grants a petition for an alternative. Supplier history by
itself is not the same as such an exemption.

Why
representative sampling matters more than a large test menu

Laboratories test the sample they receive, not the unseen remainder
of a pallet, drum, blender or warehouse.

A single scoop from the top of one container can miss segregation,
moisture pockets, localized contamination or container-to-container
differences. A defensible sampling plan defines:

• the lot or batch boundary;
• the number and location of increments;
• how containers or units are selected;
• whether increments are tested separately or combined;
• the tools and cleaning controls used;
• who sampled and when;
• sample containers, seals and custody records;
• how much material is sent to the laboratory;
• how reserve samples are retained; and
• what to do when the lot is heterogeneous or the result is near a
  limit.

Composite samples can estimate an average, but they may dilute a
localized high result. Separate-unit testing can reveal variability but
costs more and still depends on selection. Neither is universally
superior. The plan should match the hazard and decision.

More frequent testing of poorly collected samples can create a false
sense of certainty. Better sampling and fewer well-justified analyses
may sometimes provide stronger information than many convenience
samples.

What “batch tested”
should mean on a website

“Batch tested” is most useful when the seller discloses enough
information for a customer to verify the claim. A strong presentation
includes:

• the finished product name and form;
• the lot or batch code shown on the package;
• the date the sample was collected and received;
• the laboratory and report identifier;
• the tested matrix, such as powder, capsule fill or extract;
• the analytes included in each panel;
• methods or method references;
• units, specifications and reporting limits;
• the actual results, including how ND, LOD and LOQ are used;
• a clear pass/fail or quality disposition where appropriate;
• any qualifications or excluded tests; and
• a link that matches the current lot rather than a generic
  example.

The claim should not imply that:

• the laboratory sampled the lot unless it actually did;
• every package was individually tested;
• every possible contaminant was screened;
• passing a company specification equals FDA approval;
• an ISO/IEC 17025-accredited laboratory means every listed method is
  within its accredited scope;
• one historical COA applies to future lots; or
• a passing result in one jurisdiction guarantees legal compliance
  everywhere.

Plain language earns more trust than an oversized “LAB TESTED”
badge.

A
12-step method for building or reviewing a testing schedule

Step 1: Map every
material and product type

List incoming leaf, powder, capsule shells, extracts, flavorings,
packaging components and finished products. Keep ordinary botanical leaf
separate from extracts, enhanced material, concentrated 7-OH and
manufactured derivatives.

Step 2: Define batches and
lots

Document what creates a new incoming lot, manufacturing batch,
packaging lot and finished-product lot. Make the lot code traceable
through the full record.

Step 3:
Write specifications before reviewing results

For each material and product, state identity, composition,
contamination and packaging specifications. Do not choose a pass limit
after seeing the result.

Step 4: Identify
each hazard and control stage

Decide whether a risk is best controlled at supplier qualification,
receipt, processing, finished release, packaging, storage or stability
monitoring.

Step 5: Select
fit-for-purpose methods

Verify that the method is suitable for the matrix and concentration
range. Confirm analytes, extraction, calibration, reporting limits and
units.

Step 6: Design
representative sampling

Define the lot population, locations, number of increments, composite
or discrete approach, custody controls and reserve quantity.

Step 7: Assign a base
frequency

State whether each control occurs for every incoming lot, every
process batch, each packaged lot, a statistically selected subset,
planned stability time points or only after defined triggers.

Step 8: Document the
rationale

Connect the frequency to hazard severity, variability, historical
data, supplier reliability, process capability and legal
consequences.

Step 9: Establish
escalation triggers

Write down what causes intensified testing, a hold, a second
laboratory, supplier requalification or a broadened panel.

Step 10: Establish
reduction criteria

If reduced testing is possible, specify the evidence required,
approval authority, duration and automatic return-to-full-testing
conditions.

Step 11: Review results
before release

Confirm lot identity, sample traceability, laboratory authenticity,
units, specifications, deviations and any near-limit uncertainty before
disposition.

Step 12: Trend and review
the program

At a planned interval, evaluate failures, near limits, supplier
performance, complaints, regulatory changes and whether the sampling
plan still answers the intended questions.

Five fictional
testing-program reviews

These examples are educational and do not describe Kiody products or
suppliers.

Example 1: A new botanical
supplier

Claim: “The supplier tests every harvest, so we will
accept its COA.”

Review: The buyer has not confirmed the supplier’s
results, authenticated the laboratory or checked whether the pesticide
panel reflects the growing region. The first shipments also show wider
moisture variation than the current supplier.

Better approach: Hold each incoming lot; conduct
identity verification and risk-based confirmation testing; compare
results with the supplier COA; qualify the supplier only after
documented evidence; set a periodic reconfirmation schedule and
escalation rules.

Example 2: Plain
powder converted to capsules

Claim: “The powder passed, so the capsule lot needs
no further checks.”

Review: The powder result may remain relevant, but
encapsulation introduces fill-weight, line-clearance, shell and
packaging questions. The connection between the powder lot and capsule
batch is not documented.

Better approach: Preserve lot genealogy, monitor
fill weights, verify labels and seals, assess whether selected
finished-lot testing is needed, and link the public COA to the capsule
lot—not merely to a generic powder name.

Example 3: A standardized
extract

Claim: “We test residual solvents once a year
because the extraction recipe is unchanged.”

Review: Calendar frequency alone does not address
batch variation, solvent recovery, equipment changes or a failed drying
step. The product’s extract concentration increases the consequence of
composition errors.

Better approach: Tie residual-solvent verification
to each relevant extraction or finished lot unless robust validation and
a documented alternative control justify another schedule. Add triggers
for equipment, solvent, temperature, recovery or supplier changes.

Example 4: A
near-threshold 7-OH result

Claim: “The COA says 0.04%, so it passes
everywhere.”

Review: The result is meaningless for
multijurisdictional legal review without confirming the denominator,
dry-weight correction, serving amount, uncertainty and applicable state
or local rule. Some rules use a different unit or prohibit the category
outright.

Better approach: Hold the lot for legal and quality
review, confirm units and method performance, evaluate measurement
uncertainty and sampling representativeness, and do not generalize one
threshold across jurisdictions.

Example
5: A long-qualified supplier with an upward lead trend

Claim: “All results still pass, so the reduced
testing schedule can continue.”

Review: Individual passes can conceal a meaningful
trend. The last five results are moving steadily toward the internal
limit.

Better approach: Intensify confirmation testing,
investigate origin and processing changes, review sampling and
analytical consistency, and reassess the supplier before an actual
specification failure occurs.

Twenty
warning signs in a testing-frequency claim

1. “Lab tested” appears without a lot-linked report.
2. “Batch tested” does not define what a batch is.
3. The same COA appears beside multiple unrelated products.
4. The report date predates the represented lot without an explained
   connection.
5. The product form on the report differs from the product sold.
6. Only one analyte was measured, but the page implies a full safety
   panel.
7. A supplier COA is presented as independent verification by the
   seller.
8. The laboratory name or report number cannot be authenticated.
9. The sampling party and sample source are undisclosed.
10. The method is unsuited to the matrix or target range.
11. Reporting limits are higher than the claimed specification.
12. Units or denominators are missing.
13. ND is translated as “zero” or “completely free.”
14. A percentage result is compared with a milligram-per-serving limit
    without conversion data.
15. Testing frequency never changes after supplier or process
    changes.
16. Failed or near-limit results do not trigger investigation.
17. A passing release COA is used as proof of shelf life.
18. ISO/IEC 17025 accreditation is described as product approval.
19. The program has no reserve-sample or chain-of-custody records.
20. “FDA compliant” is used to imply that FDA approved a kratom
    product.

Testing-frequency record
template

For each material, process or finished product, a review record can
capture:

1. Product or material name.
2. Internal item code.
3. Product category.
4. Incoming supplier.
5. Manufacturer or processor.
6. Country and region of origin, when known.
7. Incoming lot definition.
8. Manufacturing batch definition.
9. Packaging-lot definition.
10. Finished lot code.
11. Intended market jurisdictions.
12. Applicable age restriction.
13. Applicable product-category restrictions.
14. Applicable 7-OH thresholds and denominators.
15. Controlled-derivative exclusion controls.
16. Identity specification.
17. Composition specifications.
18. Microbial specifications.
19. Elemental-contaminant specifications.
20. Pesticide specifications.
21. Mycotoxin specifications.
22. Residual-solvent specifications.
23. Moisture or water-activity specifications.
24. Allergen or cross-contact controls.
25. Foreign-material controls.
26. Packaging and label specifications.
27. Test or examination method.
28. Method source and version.
29. Matrix suitability evidence.
30. Laboratory name and location.
31. Accreditation and scope review date.
32. Sampling party.
33. Sampling plan identifier.
34. Number and location of increments.
35. Composite or discrete sample decision.
36. Chain-of-custody record.
37. Reserve-sample quantity and location.
38. Base testing frequency.
39. Frequency rationale.
40. Supplier-COA reliance, if any.
41. Initial confirmation evidence.
42. Periodic reconfirmation interval and rationale.
43. Finished-batch selection method.
44. Stability-study protocol and time points.
45. Escalation triggers.
46. Reduction criteria.
47. Most recent result and disposition.
48. Trend-review outcome.
49. Quality approver and approval date.
50. Next scheduled program review.

The record should point to controlled documents rather than
duplicating changing procedures in several places.

Frequently asked questions

How often should kratom be
tested?

There is no universal calendar interval. Frequency should be defined
separately for incoming identity, contaminants, in-process controls,
finished-batch verification, packaging, stability and investigations.
Lot variability, product type, supplier history, process controls and
applicable law all matter.

Does “batch
tested” mean every package was tested?

No. A laboratory usually tests a sample intended to represent a batch
or lot. The sampling plan determines how well the sample represents the
untested remainder.

Does
every finished batch need a full laboratory panel?

Not necessarily. A quality system may place controls at receipt,
during processing and at finished release, and may use a statistically
selected subset for some finished-batch verification. The plan needs
written scientific and quality justification.

Should
every incoming leaf lot receive identity testing?

Under the Part 111 quality benchmark, each unique lot of a dietary
ingredient generally receives identity testing unless FDA grants an
alternative through the petition process. FDA’s separate position is
that kratom is not lawfully marketed as a dietary supplement.

Can a seller rely on a
supplier COA?

For some non-identity component specifications under Part 111,
reliance can be part of a controlled program when the supplier is
qualified, results are initially confirmed, the COA contains necessary
information, qualification is documented and reliability is periodically
reconfirmed. A supplier logo or long relationship alone is
insufficient.

What does “periodically
tested” mean?

The phrase has no useful precision by itself. A program should state
the event or interval, population, tests selected, justification and
triggers that increase frequency.

Is annual testing enough?

An annual calendar event cannot represent every new lot or answer
every risk. It may be part of supplier reconfirmation, method review or
stability work, but it is not a universal substitute for lot-linked
controls.

Is monthly
testing better than quarterly testing?

Not automatically. The useful comparison is whether the plan samples
the right material, at the right process stage, with a suitable method
and sufficient frequency to detect meaningful variation.

Does a large panel
make a COA more reliable?

Not by itself. Panel size does not prove representative sampling,
valid methods, suitable reporting limits, correct lot identity or
accurate interpretation.

Who should collect the
sample?

An independent sampler can reduce some conflicts, but manufacturer-
or supplier-collected samples can still be informative when a controlled
representative procedure and chain of custody are documented. The report
should not imply laboratory sampling if the laboratory only received a
submitted sample.

What is a reserve sample?

A reserve sample is retained material from a distributed lot that can
support later examination or investigation. Under the Part 111
benchmark, reserve samples are retained in at least twice the quantity
needed for all required tests or examinations, subject to the
regulation’s conditions.

Should
plain leaf and extracts have the same schedule?

No assumption should be made that they do. Extraction can change
composition, concentrate constituents and introduce residual-solvent or
process-control questions. The schedule should reflect the actual matrix
and process.

Does a
powder COA cover capsules made from that powder?

Only if records connect the exact powder lot to the capsule batch and
the quality system addresses risks added during encapsulation and
packaging. The public presentation should make that connection
clear.

Does an
alkaloid test prove a product is botanical leaf?

No. Alkaloid analysis measures reported compounds. Botanical identity
may require other evidence, and a composition result alone does not
distinguish all possible sources or manufacturing histories.

How often should
microbial testing occur?

It should follow a written risk assessment covering incoming
contamination, any validated reduction step, post-treatment exposure,
sanitation, storage and historical results. A wet shipment, sanitation
deviation or adverse trend should trigger additional evaluation.

How often should heavy
metals be tested?

Frequency should reflect origin, soil and supply-chain risk, supplier
data, historical variability and product specifications. A historical
result for one lot does not establish the content of all later lots.

Does a release COA
prove the expiration date?

No. Release testing describes a sample at release. Shelf-life claims
require stability evidence tied to the product, packaging and storage
conditions.

What happens after a
failing result?

The lot should be controlled while the result, method, sample,
calculation and manufacturing record are investigated. Testing into
compliance—repeating tests until a passing result appears without a
scientifically justified investigation—is not a sound practice.

Should
a near-limit result be treated like an ordinary pass?

Not necessarily. A near-limit result can justify review of
measurement uncertainty, rounding, sampling variability, trends and
jurisdiction-specific decision rules. It may warrant a hold or
confirmation even when the reported number is technically within an
internal limit.

Can one 7-OH
result establish legality nationwide?

No. Jurisdictions can use different definitions, percentages,
dry-weight bases, total-alkaloid denominators, milligrams-per-serving
caps or categorical bans. Laws also change. The destination and current
official source must be reviewed.

Is
ordinary botanical leaf the same as concentrated 7-OH?

No. Ordinary leaf naturally contains a complex alkaloid profile and
is not equivalent to a concentrated, fortified, synthesized or
semi-synthesized 7-OH product. Product form, manufacturing history and
quantitative testing matter. Kiody does not sell concentrated 7-OH.

Does DOJ’s
MGPI policy create a legal trace limit?

No. The September 1, 2026 statement describes enforcement discretion
for incidental trace MGPI in otherwise botanical kratom. It does not
remove MGPI from Schedule I and provides no numerical safe harbor.

Does laboratory
testing prove a product is safe?

No test can establish zero risk. Testing provides evidence about the
sampled material, listed analytes, method and reporting limits. It
should be part of a broader quality system that includes supplier
control, manufacturing records, packaging, complaints, adverse-event
handling and recalls.

Is an ISO/IEC 17025
laboratory enough?

Accreditation is relevant, but the certificate, location and
technical scope should be checked. A laboratory can be accredited
without every method or matrix on a kratom COA appearing within its
accredited scope. Accreditation also does not mean FDA approved the
product.

Primary sources

1. U.S. Food and Drug Administration, FDA
   and Kratom (current federal agency position; accessed September 3,
   2026).
2. Electronic Code of Federal Regulations, 21
   CFR Part 111—Current Good Manufacturing Practice in Manufacturing,
   Packaging, Labeling, or Holding Operations for Dietary Supplements
   (accessed September 3, 2026).
3. U.S. Food and Drug Administration, Small
   Entity Compliance Guide: Current Good Manufacturing Practice in
   Manufacturing, Packaging, Labeling, or Holding Operations for Dietary
   Supplements (December 2010; accessed September 3, 2026).
4. U.S. Food and Drug Administration, Investigations
   Operations Manual, Chapter 4—Sampling (2026 edition; accessed
   September 3, 2026).
5. U.S. Food and Drug Administration, Guidelines for the
   Validation of Chemical Methods for the FDA Foods Program
   (method-validation reference; accessed September 3, 2026).
6. Department of Health and Human Services, 7-Hydroxymitragynine
   Above a Specified Threshold in Schedule I; Extension of Comment
   Period (August 26, 2026).
7. Drug Enforcement Administration, Temporary
   Placement of Mitragynine Pseudoindoxyl, MGM-15, and MGM-16 in Schedule
   I (effective August 26, 2026).
8. U.S. Department of Justice, Justice
   Department Announces Emergency Scheduling of Three Potent Opioid
   Compounds (updated September 1, 2026).