What are reserve (retention) samples, how much must be kept, and for how long?

1Reference, retention and reserve samples.

The US drug CGMP regulations use one term, the reserve sample. 21 CFR 211.170 requires “an appropriately identified reserve sample that is representative” of each lot in each shipment of each active ingredient, and of each lot or batch of drug product.¹

EU GMP Annex 19 separates two purposes. A reference sample is a sample of a batch of starting material, packaging material or finished product “stored for the purpose of being analysed should the need arise”. A retention sample is a fully packaged unit of finished product, kept for identification: presentation, packaging, labelling, leaflet, batch number and expiry date. For finished products the two are often identical and “may be regarded as interchangeable”.² The samples serve as a record of the batch that can be assessed in the event of, for example, a quality complaint, a query about compliance with the marketing authorisation, a labelling or packaging query, or a pharmacovigilance report.²

For active pharmaceutical ingredients, ICH Q7 says reserve samples are for “potential future evaluation of the quality of batches of API and not for future stability testing purposes”.³

2How much to keep and for how long.

The US periods differ for radioactive drug products (3 months after expiry if the dating period is 30 days or less, otherwise 6 months) and for OTC products exempt from expiry dating (3 years after distribution).¹ In the EU, a longer period may be required by the law of the Member State of manufacture, and national requirements on sample size apply where they exist.² The EU GMP glossary defines a starting material as “Any substance used in the production of a medicinal product, but excluding packaging materials”, so excipients are included.⁴

ICH Q7 sets the API manufacturer’s own rule: keep reserve samples of each batch for 1 year after the expiry date assigned by the manufacturer, or 3 years after distribution of the batch, whichever is longer. For an API with a retest date, keep them for 3 years after the batch is completely distributed. The quantity should allow at least two full compendial analyses, or two full specification analyses where there is no monograph.³

3Storage, inspection and access.

US drug product reserve samples are stored under conditions consistent with the labelling. Except for radioactive drug products, reserve samples from representative lots, selected by acceptable statistical procedures, are examined visually at least once a year for deterioration unless that would affect the sample’s integrity. Any deterioration is investigated under 21 CFR 211.192, and the results are kept with the product’s stability data.¹

Annex 19 adds expectations about where samples are and who can reach them.²

• Storage follows the marketing authorisation, for example refrigeration where relevant.
• Where a batch is packaged in two or more distinct operations, at least one retention sample is taken from each.
• The analytical materials and equipment for the specification tests stay available, or readily obtainable, until one year after expiry of the last batch manufactured.
• The Qualified Person who certifies a batch should ensure its samples are accessible at all reasonable times, with written agreements between sites where more than one is involved.
• Retention samples are kept within the EEA, preferably where the certifying Qualified Person is located.
• If a manufacturer closes, its samples move to an authorised storage site under arrangements the competent authority accepts.

The EMA’s GMP questions and answers settle two common points. A sterile product’s retained samples need not hold the full number of units for repeating the release sterility test, only enough for a confirmatory test on two occasions using the minimum quantities in the European Pharmacopoeia. The supervisory authority can grant, on request, an exemption from keeping a fully packaged unit where criteria such as a batch of fewer than 50 units or a high-value, low-volume product are met.⁵

4What the 2026 revision of Annex 19 changed.

The Commission adopted a revised Annex 19 as C(2026) 4135 on 23 June 2026, and EudraLex Volume 4 lists it as applicable as of 24 September 2026.²⁶ The stated reason is parallel imported, distributed and traded products, and the rewritten section 9 sets out:²

• packaging materials used in re-packaging are kept for the shelf life of the re-packaged product;
• reference samples of the re-packaged product are not required;
• a retention sample of the re-packaged product, with its primary and secondary package, is taken for each re-packaging operation and kept at least one year after expiry, though where the secondary package is not opened only the packaging material used needs to be kept;
• where a physical retention sample “cannot reasonably be retained” and the competent authority has agreed in advance, a photographic or digital sample may be kept instead. It should be a complete record of both packages, show the safety features and braille, trace to the batch packaging record, and follow the principles of Annex 11 if stored electronically.

The retention periods and sample sizes for manufacturers are the same as in the 2006 version, which came into operation on 1 June 2006.⁷

5Bioavailability and bioequivalence samples.

Separate US rules apply to samples from bioavailability and bioequivalence studies. Under 21 CFR 320.38, the applicant or contract research organisation keeps reserve samples of the test article and, for bioequivalence comparisons, the reference standard, enough for FDA to perform all the release tests in the application five times. They are stored under labelled conditions, in an area segregated from testing with limited access, for at least 5 years after the application is approved, or 5 years after the study is completed if it is not.⁸ Section 320.63 applies the same retention rules to in vivo and in vitro bioequivalence studies for abbreviated applications.⁹

6How it runs in Seal.

In Seal’s sample management, retention follows the sample’s governing study, batch, protocol or commitment, and an active hold blocks disposal even after the retention date has passed. Each custody event keeps who moved the sample, when, from where to where, its quantity and condition, and disposal records its scope, authorisation, method and any required witness. A contract testing laboratory can calculate the quantity a test panel needs and protect retain and repeat-test reserves, so an insufficient sample prompts an accountable decision rather than silent overconsumption.

When a result needs a repeat, Neil, Seal’s AI agent, can check sample availability for retesting against the actual aliquots, reservations and restrictions. The laboratory owner authorises the repeat and any allocation.