Standardised Herbal Extract Testing: HPLC Marker Assays and DNA Authentication (2026)

What standardised herbal extract testing has to prove

A label that says “standardised” is a promise that needs two proofs: the right plant and the right strength. Standardised herbal extract testing supplies both. A perfect marker assay on the wrong species fails the customer, and so does the right species at the wrong strength. This guide explains the chemical and DNA methods, what each can and cannot show, and how to build them into a specification.

Standardised, quantified and other extracts

The European Medicines Agency’s quality guideline for herbal medicinal products, a widely used benchmark, separates three types of extract.^[1] A standardised extract is adjusted to a defined content of one or more constituents with known therapeutic activity, using inert excipients or by blending batches. A quantified extract is adjusted to one or more active markers within a limited, specified range, by blending batches only. For other extracts, which are not adjusted, one or more analytical markers are measured batch by batch for control purposes.^[1]

So “standardised” is a specific claim. It means the extract carries a number you can test, not just a plant name.

Why the tolerance matters

For a standardised preparation, the guideline asks for the content of the active constituents to be stated with the lowest possible tolerance, with upper and lower limits (for example x% ± y%).^[1] In the finished product at release, the content is generally expected to be within ±5% of the declared value. Where the active constituents are not known and markers are used instead, a widening to ±10% can be acceptable if justified.^[1]

These limits come from EU medicines law. For an Indian health supplement under FSSAI they are design targets rather than legal limits, but they turn batch-to-batch consistency into something you can measure.

HPLC marker assays: proving strength

How the assay works

High-performance liquid chromatography (HPLC) separates the components of an extract on a column, detects the marker as a peak and quantifies it against a reference standard. The result is reported as marker content per gram of extract and compared with the acceptance range in the specification. The EMA guideline expects the analytical procedure, with details of its reference standards, to be described together with the limits applied.^[1] Our guide to analytical techniques explains how HPLC compares with other instrumental methods.

What makes the number trustworthy

An HPLC result is only as good as its validation. ICH Q2(R2) requires an analytical procedure to be shown fit for its intended purpose.^[2] For a marker assay, that means showing specificity (the marker peak is free of interference), accuracy and precision over the reportable range. To assess linearity, the guideline recommends a minimum of five concentrations spread across the range.^[2] Where one procedure lacks specificity, it should be backed by one or more supporting procedures.^[2]

The acceptance limits themselves should be specified and justified on the basis of the validated analytical range.^[1] A marker percentage with no validated method behind it is just a number.

When one marker is not enough

A single peak can stay on target while the rest of the extract changes. Where an extract has no constituents of known therapeutic activity, the guideline says that following the markers alone will not show its stability. A series of stability-indicating tests such as TLC or HPLC is needed, with fingerprint chromatograms showing that the proportions of constituents stay comparable to the initial profile.^[1] For how markers are chosen in the first place, see our note on marker substances in herbal preparations.

HPTLC fingerprinting: a fast identity screen

High-performance thin-layer chromatography (HPTLC) runs an extract beside a reference on the same plate and compares the pattern of bands. It answers “does this look like the right herb?” quickly, and with a validated method it can also be used as an assay. Where no pharmacopoeial monograph exists, the EMA guideline expects the specification of both herbal substances and herbal preparations to use chromatographic fingerprinting.^[1]

A 2024 study of 51 processed ingredients, sold as traditional Chinese medicine decoction pieces, shows how well it screens. The authors scored band intensity from 0 to 5 against the fingerprints of pharmacopoeial botanical reference standards. Of the 50 samples analysed by HPTLC, 49 gave a positive identification of the expected ingredient.^[3] HPTLC also worked on samples from which no usable DNA could be recovered, which is why the authors placed it first.^[3]

A fingerprint confirms that the expected herb is there. It is not designed to find a small amount of something else, as our analysis of a tainted supplement recall and identity testing shows.

DNA authentication: proving the species

How DNA barcoding works

DNA barcoding reads a short, standardised region of the genome and matches the sequence against a reference database to identify the species. The Indian market study below used the chloroplast rbcL region;^[4] the 2024 study used the nuclear ITS1 and ITS2 regions.^[3] DNA metabarcoding uses high-throughput sequencing to read many sequences from one sample, so it can reveal several species at once, including substitutes and contaminants.^[3]

What Indian market samples show

Balaji and Parani tested 107 single-drug herbal powders, covering 65 species, bought in markets in Tamil Nadu. Using the rbcL barcode, they found 58 (54%) authentic and 49 (46%) adulterated.^[4] In about 29 of the adulterated samples the labelled species had been completely replaced, mostly by taxonomically unrelated plants; about 20 were mixtures.^[4] The rate varied by herb: 15 of 17 tulsi (Ocimum tenuiflorum) powders (88%) were authentic.^[4]

The samples came from one state’s markets, so they show that substitution happens, not how often it happens across India. Adulteration is not only an Indian problem either: see our summary of the BAPP bulletin on echinacea adulteration.

Where DNA testing has limits

Processing damages DNA. In the 2024 study, conventional barcoding gave only 16 positive identifications across the 51 processed ingredients, against 33 for metabarcoding, and the authors attribute the poor results largely to DNA degradation.^[3] Gaps in reference databases also limit what barcoding can identify.^[3] DNA confirms the species, not the strength, so it cannot replace chemistry.

For an extract, ask whether identity was tested on the raw herb before extraction, when intact DNA is most likely to be present (FRL view).

Combining chemical and DNA methods

No single test proves a standardised herbal extract. The 2024 authors propose a tiered strategy. HPTLC comes first, because it screens every sample, including those with no usable DNA. DNA barcoding comes second, to confirm the target species, and metabarcoding third, to show what else is present.^[3] Add a validated quantitative assay, usually HPLC, for strength.^[1][2]

Table 1. What each method proves

Formulator’s note. A certificate that gives a marker percentage without naming the method, the reference standard and its validation status cannot be compared between batches or between suppliers.

Writing the specification and questioning your supplier

Under the EMA guideline, a herbal substance that has a pharmacopoeial monograph must comply with it, and a herbal preparation generally should. Where there is no monograph, the specification should cover characteristics, identification, assay and purity, with tests for microbial quality, mycotoxins, pesticide and fumigant residues, heavy metals and adulterants.^[1] For Indian supply, check whether the ingredient has a monograph in the Ayurvedic Pharmacopoeia of India or the Indian Pharmacopoeia, and write its identity and assay tests into your specification. For quality requirements on the AYUSH route, see our AYUSH guidelines for herbal products.

Put these questions to every supplier before the first pilot batch.

Table 2. Supplier questions for a standardised extract

On stability, the guideline’s shelf-life benchmark for marker content is ±5% of the initial batch value, or up to ±10% if justified.^[1] An accelerated study can give an early estimate. In one Indian example, a herbal food supplement held at 40 °C and 75% relative humidity was tested at 0, 2, 4 and 6 months, and a shelf life of about 24 months was extrapolated using a 10% degradation threshold.^[5] Real-time data should then confirm it.

In practice: testing on a real product file

Our ashwagandha capsule development worked example applies these tests to one product: HPLC or HPTLC for strength and DNA barcoding for species, with both in the file before the first batch. It also shows why the extract itself matters. The 240 mg dose in the Lopresti 2019 trial belongs to one extract, Shoden, standardised by HPLC to at least 35% withanolide glycosides.^[6] Our Lopresti 2019 ashwagandha study explainer covers that trial in detail.

From the FRL lab

In an FRL Inside the Lab study of a purple tea beverage over 90 days of storage, total polyphenols (ISO 14502-1) fell by about 5–6% and anthocyanins by about 12–13%, while both variants kept acceptable microbiological characteristics.^[7] Summary assays like these show that something is changing, not which compound. The team’s recommended next step was LC-MS/HPLC profiling to quantify individual anthocyanins and catechins.^[7]

The beverage is a plant-extract product, not a standardised extract, but the lesson carries over: a total-content figure is not a marker assay.

Conclusion

A credible standardised extract is proven twice. Chemistry shows the strength; fingerprinting and DNA show the species. Both need validated methods and stated acceptance limits. Ask every supplier for the marker, the method, the reference standard and the identity evidence behind the certificate, and test incoming lots against your own specification. For the wider path from extract choice to stability and route, see our herbal product development process guide.

For brands & formulators

Food Research Lab’s herbal formulation development team can help you choose markers, write an extract specification with identity, assay and contaminant limits, and plan HPLC or HPTLC assays and DNA identity checks, which we run or arrange through accredited partner laboratories. We can also review a supplier’s certificates of analysis before you commit to a pilot batch.

Once the specification is settled, our herbal manufacturing team supports pilot and contract batches, and our regulatory and scientific research team checks the route and label. See also our herbal product development sector page, or contact us with your brief.

In this series

This page is part of FRL’s herbal and Ayurvedic product development series for India.

• Herbal product development process: the pillar guide, seven stages from plant to shelf.
• The ashwagandha stress study (Lopresti 2019) explained: what one trial does and does not support.
• AYUSH licence vs FSSAI: which regulatory route a herbal product takes in India.
• Herbal product development in India: 10 questions: short answers on extracts, testing, stability and regulation.
• Ashwagandha capsule development: a worked example: one product from brief to stability-ready specification.
• Standardised herbal extract testing: HPLC and DNA: You are here. How strength and identity are proven.
• Private label herbal supplements in India: costs, MOQs and lead times for brands.
• Service: herbal formulation development.

1. European Medicines Agency, Committee on Herbal Medicinal Products. (2022). Guideline on quality of herbal medicinal products/traditional herbal medicinal products (EMA/HMPC/CHMP/CVMP/201116/2005 Rev. 3). https://www.ema.europa.eu/en/documents/scientific-guideline/final-guideline-quality-herbal-medicinal-productstraditional-herbal-medicinal-products-revision-3_en.pdf (cited for standardised, quantified and other extracts; the x% ± y% tolerance; ±5% release and shelf-life limits with widening to ±10% if justified; analytical procedures and reference standards; acceptance limits based on the validated range; stability-indicating TLC/HPLC fingerprints; monograph compliance; comprehensive specifications with fingerprinting and contaminant tests).
2. International Council for Harmonisation. (2023). ICH Q2(R2) guideline on validation of analytical procedures (EMA/CHMP/ICH/82072/2006, Step 5). European Medicines Agency. https://www.ema.europa.eu/en/documents/scientific-guideline/ich-q2r2-guideline-validation-analytical-procedures-step-5-revision-2_en.pdf (cited for fitness for intended purpose, specificity, accuracy and precision over the reportable range, a minimum of five concentrations for linearity, and supporting procedures where specificity is lacking).
3. Mück, F., Scotti, F., Mauvisseau, Q., Thorbek, B. L. G., Wangensteen, H., & de Boer, H. J. (2024). Three-tiered authentication of herbal traditional Chinese medicine ingredients used in women’s health provides progressive qualitative and quantitative insight. Frontiers in Pharmacology, 15, 1353434. https://doi.org/10.3389/fphar.2024.1353434 (cited for the 51 processed decoction-piece samples, 0–5 band intensity scoring against pharmacopoeial botanical reference standards, 49 HPTLC identifications of 50 analysed, 33 metabarcoding and 16 barcoding identifications, nrITS1/nrITS2 markers, DNA degradation, reference-database gaps and the tiered strategy).
4. Balaji, R., & Parani, M. (2022). DNA barcoding of the market samples of single-drug herbal powders reveals adulteration with taxonomically unrelated plant species. Diversity, 14(6), 495. https://doi.org/10.3390/d14060495 (cited for 107 powders from 65 species in Tamil Nadu markets, rbcL barcoding, 58 authentic and 49 (46%) adulterated, the substitution and mixture shares, and 15 of 17 tulsi powders authentic).
5. Dorni, A. I. C., Venkatesan, U., Srinivasan, R., Moorthy, S., & Raja, K. (2023). Evaluation of stability and determination of shelf life of an herbal food supplement: Naturovita. Journal of Pharmacognosy and Phytochemistry, 12(1), 219–225. https://doi.org/10.22271/phyto.2023.v12.i1c.14580 (cited for the 40 °C/75% RH accelerated design, testing at 0, 2, 4 and 6 months and the 24.08-month estimate from a 10% degradation threshold).
6. Lopresti, A. L., Smith, S. J., Malvi, H., & Kodgule, R. (2019). An investigation into the stress-relieving and pharmacological actions of an ashwagandha (Withania somnifera) extract: A randomized, double-blind, placebo-controlled study. Medicine, 98(37), e17186. https://doi.org/10.1097/MD.0000000000017186 (cited for the 240 mg Shoden extract standardised by HPLC to not less than 35% withanolide glycosides).
7. Food Research Lab. (2026, August 26). Functional characterization and shelf-life evaluation of purple tea beverage [Inside the Lab]. https://www.foodresearchlab.com/insights/inside-the-lab/purple-tea-beverage/ (cited for the ISO 14502-1 and DPPH methods, the 90-day polyphenol and anthocyanin results, microbiological stability and the recommended LC-MS/HPLC profiling).