Identity Testing 111 75
21 CFR 111.75(a)(1)(i) requires the supplement manufacturer to conduct at least one appropriate test or examination to verify the identity of every incoming lot of every dietary ingredient before it is used. Supplier CoA is not a substitute. This single subsection drives more FDA 483 observations and Warning Letters in the dietary-supplement industry than any other.
01What 111.75(a)(1)(i) actually says
21 CFR 111.75(a)(1)(i) is one sentence with enormous operational weight: 'You must conduct at least one appropriate test or examination to verify the identity of any component that is a dietary ingredient, unless you petition the agency under paragraph (a)(1)(ii) of this section and the agency grants an exemption.' Every lot. Every dietary ingredient. Before release. By the supplement manufacturer themselves — not by relying on the supplier's CoA. This single subsection is the most-cited cGMP failure in FDA Warning Letters issued under 21 CFR Part 111, present in roughly 60-70% of all supplement Warning Letters across any given year.
02What counts as a 'dietary ingredient'
The rule applies specifically to components that are dietary ingredients (vitamins, minerals, herbs or other botanicals, amino acids, dietary substances used to supplement the diet, concentrates / metabolites / constituents / extracts / combinations thereof). Excipients (binders, fillers, disintegrants, lubricants, coatings, encapsulant gelatin or HPMC) are NOT dietary ingredients and are NOT required to be identity-tested under (a)(1)(i) — they fall under §111.75(a)(2), which permits qualified-supplier reliance with a documented qualification program. The line between 'dietary ingredient' and 'excipient' matters: rice flour added as a bulking agent in a tablet is an excipient; rice flour added as a labelled source of dietary fibre is a dietary ingredient and must be identity-tested every lot.
03Why the supplier's CoA is not enough
FDA's position, restated in countless Warning Letters, is that identity verification is the manufacturer's responsibility — not the supplier's. The supplier's CoA is documentary evidence of what the supplier says they tested. It is not evidence that the material in this drum, on this day, matches that CoA. Drums get mis-labelled. Pallets get cross-shipped. Brokers re-pack. Adulterants get added (the 2013 botanical industry investigations found multiple cases of substitution and dilution). FDA's reasoning is simple: the failure mode is real, the consequence is consumer harm, and the only reliable defence is for the receiving manufacturer to test the actual lot they are about to put into product.
04What counts as 'appropriate'
'Appropriate' means scientifically valid and specific enough to distinguish the dietary ingredient from likely adulterants and substitutes. The method must be selected based on the ingredient and the realistic adulteration risk — there is no universal identity test. Common choices by ingredient class:
- Synthetic vitamins and minerals — FTIR (Fourier-transform infrared) spectroscopy is the workhorse, often with library matching. UV-Vis for some chromogenic compounds. HPLC for vitamin D, vitamin K and similar low-level analytes.
- Amino acids — typically HPLC with derivatisation, or specific rotation for chiral confirmation.
- Botanicals (whole herb, powder, extract) — macroscopic and microscopic examination per the relevant USP / Ph. Eur. monograph; HPTLC (high-performance thin-layer chromatography) fingerprint; HPLC marker-compound profile; DNA barcoding for whole-herb authentication where the genetic material is preserved.
- Probiotics — genus / species / strain ID by 16S rRNA sequencing or whole-genome sequencing; viable plate count for the strength side.
- Fish oils and omega-3 concentrates — fatty-acid profile by GC-FID.
- Proteins (whey, casein, collagen) — amino-acid profile + SDS-PAGE or HPLC for protein identity; nitrogen content alone is not identity (the 2008 melamine incident proved this).
Whatever method is chosen must be validated per §111.320 (scientifically valid method), have an associated written test procedure with acceptance criteria, and have an internal reference standard or library entry against which lot results are compared. 'We use FTIR and visually compare the spectra' is not adequate without a documented library, acceptance criteria and an analyst-qualified workflow.
05The §111.75(a)(1)(ii) petition — and why nobody qualifies
The regulation allows a manufacturer to petition FDA for an exemption from the per-lot identity-testing requirement, provided the petition demonstrates that the supplier's identity-verification system is robust enough to be relied upon. In practice almost no petitions have been granted. The bar is high (essentially, the manufacturer must show the supplier operates an equivalent QA system under the manufacturer's effective oversight, with audit rights, validated identity testing on the supplier side, and documented chain-of-custody). For 99% of supplement manufacturers, the petition pathway is theoretical and per-lot identity testing is the operational reality.
06Documentation that closes the loop
- Specification for the dietary ingredient (identity acceptance criterion + method reference) — §111.70(b).
- Written test procedure with sample prep, instrument setup, acceptance criteria, and review/approval — §111.260.
- Method validation file — accuracy, precision, specificity, limit of detection where relevant — §111.320.
- Analyst qualification record (training + competency demonstration) — §111.13.
- Reference standard or library entry traceable to USP, NIST or in-house characterised material with documented characterisation.
- Per-lot identity-test record — instrument printout, calculation, pass/fail vs the acceptance criteria, analyst sign, reviewer sign — §111.95.
- QC unit disposition of the lot — Approve (release to manufacturing) or Reject (quarantine + supplier notification) — §111.105 / §111.113.
- Linkage from the identity record to every manufactured batch record that drew from the lot — for forward traceability when an issue surfaces later — §111.260.
07Common identity-testing failures (in FDA's own words)
- No identity test at all — release based solely on supplier CoA. The single most-cited finding.
- Identity test performed but not specific — e.g. relying on Loss on Drying or appearance alone for a botanical extract.
- Method not validated — no scientific-validity evidence under §111.320.
- Analyst not qualified — no documented training or competency.
- No reference standard — visual comparison against 'previous lots' with no characterised reference.
- Identity test only on the first lot from a new supplier, not on every subsequent lot.
- QC unit signed off before the identity test was actually completed (back-dating).
- Library spectra collected once and never re-verified — instrument drift makes the library invalid.
- Botanical identified to genus but not species (e.g. 'Echinacea sp.' when the label claims 'Echinacea purpurea').
- Composite sampling across drums of a multi-drum lot without a sampling plan that proves the composite is representative.
08How this drives a buying decision
Supplement manufacturers searching for 'identity testing 111.75 software', '111.75(a)(1)(i) compliance', 'dietary ingredient identity testing SOP' or 'how to satisfy 111.75 without a lab' are almost always responding to one of three triggers: an FDA inspection that produced a 483, an upcoming third-party audit (NSF, USP, NSF Certified for Sport, UL, ISO 17025), or onboarding a new contract-manufacturer customer who is going to audit them. The buying motion is fast — the gap is concrete, the consequence is documented in public Warning Letters, and the tooling must support the receiving / sampling / testing / disposition workflow end-to-end with reviewable records.
09How V5 Ultimate handles 111.75(a)(1)(i)
- Receiving creates a quarantine lot — every incoming dietary-ingredient lot lands in QC HOLD status and cannot be consumed by a work order until the QC unit releases it.
- Identity-test workflow per ingredient — the ingredient master carries its identity test method, acceptance criteria, sampling plan and required analyst qualification; the receiving record auto-creates the identity-test task on the correct lot.
- Kiosk-driven lab capture — analyst logs in, instrument reading and result are captured against the procedure, photo of the printout is attached, and the record is signed; two-person review per §111.95 is enforced.
- Library + reference-standard register — FTIR spectra, HPLC chromatograms and microscopy reference images live in a versioned library with re-verification schedules.
- QC unit disposition (§111.105) — Approve or Reject is e-signed by a designated QC unit member; only Approved lots become consumable.
- Forward traceability — every batch record that consumed the lot is auto-linked so a later identity-test invalidation can drive recall scoping in minutes.
- Supplier scorecard wired in — repeated identity failures from one supplier surface on the supplier scorecard and trigger re-qualification.
- Industry-aware UI — supplement-industry workspaces see 'Dietary ingredient — identity test required (111.75(a)(1)(i))' inline on the receiving screen; pharma workspaces see the parallel 211.84 language instead.
Frequently asked questions
Q.Is the supplier's CoA ever enough?+
No, not for the identity test on a dietary ingredient under §111.75(a)(1)(i). The CoA is useful — it supports your specification setting, your supplier qualification under §111.75(a)(2) for non-dietary-ingredient components, and your trending — but it does not substitute for the receiving manufacturer's own per-lot identity test. The only way around the rule is the §111.75(a)(1)(ii) petition, which is rarely granted.
Q.How many drums must I sample from a multi-drum lot?+
There is no fixed FDA number. The rule requires a scientifically valid sampling plan. ANSI/ASQ Z1.4 (attribute sampling) and the √n+1 rule (test the square root of the number of containers plus one) are widely used industry practices. The key is that the plan is documented, justified, followed, and the sample is genuinely representative. Composite sampling is allowed if the plan justifies that the composite represents the lot — for high-risk botanicals subject to adulteration, per-drum testing is the safer posture.
Q.What test is appropriate for a botanical?+
It depends on the ingredient and the realistic adulteration risk. For most botanicals, a combination is used: macroscopic + microscopic examination per the relevant USP, Ph. Eur. or AHP monograph; HPTLC fingerprint compared to a reference standard; HPLC marker-compound profile; and increasingly DNA barcoding for whole-herb material. For an extract where the DNA is destroyed, chemical fingerprinting becomes the primary identity tool. The chosen combination must be specific enough to distinguish the labelled species from realistic adulterants — the famous 2015 NY Attorney General case turned on identity testing depth for store-brand herbal supplements.
Q.Does this apply to excipients too?+
No. §111.75(a)(1)(i) applies specifically to dietary ingredients. Excipients (binders, fillers, disintegrants, lubricants) fall under §111.75(a)(2), which permits qualified-supplier reliance with a documented supplier qualification program plus appropriate verification activities. The distinction is in §111.3 — if the component is a dietary ingredient, it gets the per-lot identity test; if it is purely functional and not part of the labelled dietary-ingredient content, it does not.
Q.What happens if an FDA inspector finds we are not doing identity testing?+
Expect it on the Form 483 as the lead observation. If the gap is structural (no method, no records, no QC unit decision), expect a Warning Letter under §111.75(a)(1)(i) — these are public and harm B2B contracts. The standard remediation is: stop release of affected products, set up the identity-testing program with validated methods, retroactively assess released product for safety risk, and respond in writing within 15 working days with the corrective action plan. V5 customers typically use the platform's existing receiving / QC workflow to operationalise the new program rather than building from scratch.
Primary sources
- 21 CFR 111.75 — What must you do to determine whether specifications are met?
- FDA Guidance for Industry: New Dietary Ingredient Notifications and Related Issues (2024)
- FDA Warning Letters — Dietary Supplements (searchable database)
- AOAC International — Official Methods of Analysis
- USP — Dietary Supplements Compendium and General Chapters <2030> Botanical Extracts and <2750> Manufacturing Practices
Further reading
- 21 CFR Part 111The dietary-supplement cGMP regulation identity testing lives inside.
- Scientifically valid method (111.320)The method-validation rule identity tests must meet.
- Certificate of AnalysisWhat the supplier sends — and what it does NOT replace.
- Finished product specifications (111.70(e))The downstream sibling specs identity feeds into.
- Supplier qualificationThe program identity testing supports, not replaces.
V5 Ultimate ships with the Identity Testing 111 75 controls already wired in — audit trail, e-signatures, validation evidence. Free trial, no credit card, onboard in days, not months.
