EU GMP Annex 7 for Herbal Medicinal Products: A Practical Compliance Checklist

Most EU pharmaceutical GMP auditors will tell you the same thing: herbal medicinal product manufacturers fail more often on botanical starting material documentation than anywhere else in the facility. It’s rarely the production suite that triggers a critical finding. It’s the gap between what Annex 7 of the EU GMP Guide actually demands and what a supplier’s Certificate of Analysis contains.

Annex 7 is specific to the manufacture of herbal medicinal products. While the core GMP principles in Directive 2003/94/EC apply to all medicinal product manufacturers, Annex 7 layers on requirements that reflect something unique to botanical raw materials: they’re biologically variable, highly susceptible to contamination at origin, and essentially impossible to quality-check by visual inspection alone. If you manufacture or import herbal medicinal products for the EU market — whether as licensed medicines or under the traditional herbal registration pathway established by Directive 2004/24/EC — here’s what a competent authority inspection will actually look for.

What Annex 7 Actually Covers

The current version of Annex 7 distinguishes between three separate stages in the production chain, and each carries its own documentation and testing obligations.

Herbal substances are the raw plant material — used whole, fragmented, or powdered. Think dried chamomile flower, valerian root, or echinacea aerial parts. Herbal preparations are processed forms derived from those substances: extracts, tinctures, powders from expressed juices, or distillates. Finished herbal medicinal products are the final dosage forms — tablets, capsules, oral liquids — presented for therapeutic use.

Most gaps we encounter during pre-audit gap analyses sit at the herbal substance level. Manufacturers sometimes assume that because a supplier is “approved” and delivers a COA with each batch, the incoming material is adequately characterised. Under Annex 7, that assumption doesn’t hold.

The European Pharmacopoeia (Ph. Eur.) currently contains more than 270 herbal drug monographs, and these monographs define the identity, purity, and assay requirements that apply to your specific botanical. A chamomile sourced from Egypt carries a materially different contamination profile than one from Germany or Ukraine. Annex 7 expects your testing programme to reflect that difference — not to treat all approved suppliers equally.

Step 1: Establish Botanical Identity on Every Incoming Batch

Section 7.3 of Annex 7 is unambiguous: identity must be established for the complete batch on receipt. A supplier COA confirming species does not fulfil this requirement on its own. You need your own laboratory confirmation, every time.

The acceptable methods depend on the Ph. Eur. monograph for your specific plant. Most monographs call for a combination of approaches:

• Macroscopic examination: organoleptic characteristics including colour, odour, texture, and typical morphological features
• Microscopic examination: characteristic anatomical structures such as calcium oxalate crystals, trichomes, stomata type, and endosperm cell patterns
• TLC or HPTLC fingerprinting: chromatographic profiles compared against reference standards and the monograph’s specified reference chromatogram
• DNA barcoding: increasingly used for powdered or heavily processed materials where visual methods are insufficient

DNA barcoding can confirm species identity with greater than 95% specificity for most medicinal plants, but it isn’t universally recognised in Ph. Eur. monographs as a standalone compendial method. Use it to support, not replace, pharmacopoeial identity tests.

One practical point that’s worth raising with your procurement team: if you’re receiving cut or powdered botanical material, microscopic identification becomes considerably more difficult — and sometimes unreliable. Negotiating to receive whole or coarsely fragmented material, where the supply chain allows it, preserves your identity testing options and reduces the risk of finding an ambiguous result.

Step 2: Test for Pesticide Residues — and Know Which Limits Apply

Botanical raw materials sourced from outside the EU frequently carry pesticide residue profiles that don’t align with EU expectations. The default maximum residue level (MRL) for any pesticide not listed in the annexes of Regulation (EC) No 396/2005 is 0.01 mg/kg — a stringent threshold that catches many suppliers off-guard, particularly those accustomed to domestic standards in China, India, or North Africa.

For herbal drugs specifically, Ph. Eur. general chapter 2.8.13 lists 34 pesticides with defined limits. These include chlorpyrifos (≤ 0.05 mg/kg), DDT (sum of isomers, ≤ 1.0 mg/kg), and lindane (≤ 0.6 mg/kg). For any pesticide not specifically listed, the 0.01 mg/kg default applies — and it’s your responsibility to demonstrate compliance, not the supplier’s.

If your botanical is also consumed as a food ingredient — which many popular herbs are — then Regulation 396/2005 food MRLs may apply to your raw material at the point of origin, even though your finished product is regulated as a medicine. This dual-regulation complexity catches a surprising number of otherwise well-run manufacturers who source through Asian or North African intermediaries and focus only on the pharmaceutical testing cascade.

Annex 7 doesn’t prescribe a specific testing frequency. But your quality risk management system — consistent with ICH Q10 — needs to justify whatever programme you’ve implemented. Annual testing from an established, lower-risk supplier with a controlled growing region is defensible. Testing every incoming batch from a newer supplier or an origin with a known pesticide-use intensity is expected. Inspectors will ask to see the risk assessment that underpins your decision.

Step 3: Apply the Correct Microbial Limits for Your Product Type

Ph. Eur. general chapter 5.1.8 provides tiered microbial acceptance criteria based on how the final product is prepared and consumed. Getting this wrong is a common source of minor observations — but repeated minor observations become patterns, and patterns become major findings.

The key distinctions for oral herbal products:

Product type	TAMC limit	TYMC limit
Herbal drugs for hot-water extraction (teas, infusions)	≤ 10⁷ CFU/g	≤ 10⁵ CFU/g
Other herbal drugs for oral use (not hot-water extracted)	≤ 10⁵ CFU/g	≤ 10⁴ CFU/g
Herbal preparations for oral use	≤ 10³ CFU/g	≤ 10² CFU/g

All oral herbal product categories must also demonstrate absence of Escherichia coli, Salmonella spp., Staphylococcus aureus, and Pseudomonas aeruginosa per defined sample quantities — the exact requirements vary by product category and intended route.

What inspectors check: are the microbial acceptance criteria in your product specification consistent with the Ph. Eur. limits for your specific product type and intended use? We’ve seen specifications that carried over limits from a cosmetic product formulation when the botanical moved to a medicinal application, or that applied tea-category limits to encapsulated powders that are swallowed directly. These errors look small in isolation. In an audit context, they raise questions about the adequacy of your specification review process.

Step 4: Control for Mycotoxins — Especially Aflatoxins

Aflatoxin contamination in botanical raw materials is not a theoretical risk. A 2023 survey of commercial herbal drug samples across European markets found that between 19% and 24% of lots from certain African and South Asian origins showed aflatoxin B1 levels above the Ph. Eur. 5.22 limit of 2 μg/kg. The total aflatoxin limit (B1 + B2 + G1 + G2 combined) is 4 μg/kg. Ochratoxin A has a separate limit of 20 μg/kg for affected herbal drug categories.

Your risk-based mycotoxin testing approach should account for three factors: the botanical species (some are inherently higher risk — dried figs, certain tropical plant materials, and high-starch botanicals are particularly susceptible), the geographic origin of the material, and the storage conditions throughout the supply chain. Aspergillus flavus — the primary aflatoxin-producing mould — proliferates rapidly above 70% relative humidity. If you can’t review your supplier’s humidity and temperature records for storage and container transit, that’s a supplier qualification gap, not just a testing problem.

For method selection: HPLC with fluorescence detection or LC-MS/MS are the reference methods for aflatoxin quantification. Immunoassay screening methods (ELISA) are acceptable as an initial screen but require confirmation by chromatographic analysis for any result that approaches or exceeds the limit.

Step 5: Build a Supplier Qualification File That Survives an Audit

The weakest point in most manufacturers’ Annex 7 compliance isn’t the analytical testing — it’s the traceability documentation behind the approved supplier. Section 7.2 of Annex 7 requires that each incoming batch of herbal substance be traceable to its collection or cultivation site, including:

• Geographic origin (country, region, and for wild-harvested materials, the collection area)
• Harvest or collection date and method
• Identity of the responsible person or organisation at collection
• Species identification confirmed at source — ideally by a qualified botanist or recognised herbarium reference

This is straightforward for cultivated botanicals from established European growers. It becomes considerably harder for wild-harvested material sourced through multiple trading intermediaries — which describes a significant proportion of the global trade in medicinal plants.

We routinely see supply chains where a European importer can name their Chinese or Indian exporter, but cannot confirm the province of harvest, the harvesting organisation, or whether any botanical authentication was performed at origin before processing began. That’s an untraceable supply chain by Annex 7 standards, regardless of how well-regarded the exporter is commercially.

Your approved supplier qualification file should contain: an audit report or qualified questionnaire (on-site audit preferred for higher-risk origins), a retained reference sample from initial approval, ongoing COA comparison records, and your own batch test results cross-referenced against supplier data. Any meaningful discrepancy between supplier COA values and your in-house analytical results — a concentration shift, a different pesticide profile, a change in TLC fingerprint — should trigger a formal investigation under your CAPA system. Not a note in the margin.

What EU Competent Authorities Actually Find

EU national competent authorities — ANSM in France, BfArM and PEI in Germany, AIFA in Italy — inspect herbal medicinal product manufacturers under the same GMP framework as conventional pharmaceutical manufacturers. Annex 7 doesn’t establish a more relaxed standard. If anything, the inherent variability of botanical raw materials means the quality system has to be more systematic, not less, to achieve the same level of process control that a synthetic active pharmaceutical ingredient provides by its nature.

The Traditional Herbal Medicinal Products registration pathway under Directive 2004/24/EC requires a minimum of 30 years of documented traditional use — at least 15 of those years within the EU — before a simplified application can be submitted. But GMP compliance is mandatory from the first batch you manufacture. There’s no provisional phase.

If you’re preparing for your first GMP inspection or have received critical or major observations related to botanical starting materials in a previous inspection, a structured gap analysis against Annex 7 and the Ph. Eur. monographs for your specific botanicals is the right starting point. The findings that derail inspection outcomes are almost always things that were identifiable months in advance — incomplete traceability chains, missing identity test records, microbial limits copied from the wrong product category. None of those require a laboratory rebuild to fix. They require a methodical review of what your procedures actually say versus what Annex 7 actually requires.

---

Written by Nour Abochama, Quality & Regulatory Advisor, Care Europe | VP Operations, Qalitex. Learn more about our team

Talk to our team about EU market entry and GMP readiness for herbal medicinal products. Contact us

Related from our network

• ISO 17025-accredited botanical identity and herbal raw material testing — Qalitex Laboratories offers HPTLC fingerprinting, DNA barcoding, pesticide residue panels, and mycotoxin analysis for herbal drug manufacturers requiring pharmacopoeial-grade test reports.
• Health Canada NHPD-aligned testing for herbal natural health products — Androxa supports European herbal product brands entering the Canadian market with supplier qualification testing and NHPD-compliant quality documentation.