Peptide Vial Content Uniformity: What USP 905 and Batch Sampling Say About a COA

A certificate of analysis reports what was measured in a sample drawn from a batch, not what is inside the specific vial a laboratory receives. Peptide vial content uniformity is a separate quality attribute from chromatographic purity, and it is governed by fill process control rather than by the analytical method that produced the purity number. A batch can hold a genuine 99 percent purity result while individual vials vary in how much peptide they actually contain.

This distinction is routinely collapsed in the research peptide market. Vendors publish a purity figure and treat it as a complete description of product quality. Purity is a ratio. It describes the composition of the material that was tested, expressed as the proportion of target peptide relative to everything else the detector saw. It says nothing about mass. Two vials filled from the same homogeneous bulk lot can both be 99 percent pure and still differ by a wide margin in absolute peptide content, because content is set at the filling step and purity is set at the synthesis and purification steps. These are different unit operations with different failure modes and different controls.

Key Research Findings

• Ashraf, Mackey, Vida and colleagues (Journal of Medical Internet Research, 2024, volume 26, article e65440, doi 10.2196/65440) purchased semaglutide vials from unlicensed online sellers and found the measured peptide content substantially exceeded labeled amounts by 28.56 percent to 38.69 percent.
• In the same dataset, measured purity ranged from 7.7 percent to 14.37 percent against a 99 percent label claim, and endotoxin was detected in every sample at 2.1645 to 8.9511 EU/mg.
• That study screened 1080 search engine results, identified 317 pharmacy links of which 134 led to 59 unique illegal websites, and recorded more than 4.7 million visits to the top 30 affiliated domains between July and September 2023.
• Visual inspection of the delivered vials indicated noncompliance in 59 percent to 63 percent of the evaluated packaging criteria.
• USP General Chapter 905 sets the acceptability constant k at 2.4 when 10 units are assayed and 2.0 when 30 units are assayed, with a maximum acceptance value L1 of 15.0 and an individual unit limit L2 of 25.0 unless otherwise specified.
• Hei, She, Chen, Jin, Sun, Tu and Guo (PDA Journal of Pharmaceutical Science and Technology, 2025, volume 79, issue 2, pages 157 to 169, doi 10.5731/pdajpst.2023.012867) found viscosity, target fill volume, pump tubing size, pump speed, acceleration and deceleration rate, and suck-back all exerted statistically significant influence on fill volume variability across solutions of 1 to 23 cp and fill volumes of 0.2 to 2.0 mL.

A Batch Result Is a Statistical Inference, Not a Vial Measurement

Analytical testing is destructive. The vial that goes to the laboratory is consumed by the assay, which means the vial a researcher opens has by definition never been tested. Every certificate of analysis is therefore an inference: the tested units are assumed to represent the untested ones. The strength of that inference depends entirely on how the sample was selected and how many units were drawn, and neither of those facts appears on a typical research peptide certificate.

Sampling practice in pharmaceutical manufacturing is explicit about this. Units are drawn across the fill run rather than from a single point, because fill variation is usually systematic rather than random. Pump performance drifts over a run. Tubing relaxes. Solution level in the reservoir falls and changes the head pressure feeding the pump. A sample taken entirely from the beginning of a run cannot detect a downward drift at the end of it. The commonly used square root of n plus one heuristic, where n is the number of containers, has no statistical basis and does not scale correctly across batch sizes, yet it remains widespread precisely because it is simple.

For a research buyer this produces a specific and answerable question. A certificate that identifies the batch, the testing laboratory, and the date of analysis supports a traceable claim about that batch. A certificate with no batch identifier supports nothing, because it cannot be connected to the material in hand. This is the practical reason batch-specific documentation matters more than the headline purity figure, and it is why Maple Research Labs publishes batch-linked certificates of analysis rather than a single generic document reused across production runs.

How USP General Chapter 905 Frames Unit-to-Unit Variation

USP General Chapter 905 is the harmonized standard, aligned with the European Pharmacopoeia and the Japanese Pharmacopoeia, that defines how uniformity across individual units is demonstrated. It is worth understanding because it shows what a rigorous answer to the vial variation question looks like, and by contrast how little the research peptide market currently does.

The chapter allows two routes. Content uniformity assays individual units directly. Weight variation assays a representative composite and then estimates individual unit content from individual unit weights, which is only valid when concentration is genuinely uniform across units. Under either route, 10 units are assayed first and an acceptance value is calculated from the mean, the sample standard deviation, and an acceptability constant k of 2.4. If that acceptance value is at or below 15.0, the batch conforms. If it exceeds 15.0, a further 20 units are assayed, the acceptance value is recalculated across all 30 units with k reduced to 2.0, and the batch conforms only if the recalculated value is at or below 15.0 and no individual unit falls outside 25 percent of the reference value M.

The structure of the acceptance value formula is the informative part. It penalizes two things simultaneously: deviation of the mean from the target, and dispersion around that mean. A batch centred perfectly on target but with wide unit-to-unit scatter fails. A batch that is tightly controlled but systematically offset also fails once the mean moves outside the 98.5 to 101.5 percent indifference window. Both failure modes are invisible to a purity assay.

Why Freeze-Dried Vials Fall Under Weight Variation

The chapter specifically addresses solids prepared from true solutions and freeze-dried in their final containers, and places them in the weight variation category. The logic is that when a homogeneous solution is filled and then lyophilized in the vial without further manipulation, concentration is uniform by construction and the only remaining variable is how much solution went into each container. Content variation collapses into fill variation.

This applies directly to research peptides, which are almost universally supplied as lyophilized powder in the final vial. It means the controlling parameter for vial content is the accuracy and precision of the filling equipment, not anything the purification chemistry did. It also means fill weight data, which manufacturers routinely capture in-process, is the correct evidence for content uniformity. That data almost never reaches the buyer.

What Fill Process Research Shows About the Sources of Variation

Hei and colleagues, publishing in the PDA Journal of Pharmaceutical Science and Technology in 2025, examined peristaltic pump filling using a bench-top system with a high-precision balance and evaluated performance against three criteria: filling accuracy, filling process capability, and filling precision. They tracked four statistical indexes, including relative error mean, relative standard deviation, relative moving range mean, and a process capability threshold of Cpk at or above 1.33.

Their results identified viscosity, target fill volume, pump tubing size, pump speed, acceleration and deceleration rate, and suck-back as factors with statistically significant influence on fill volume variability. They applied Definitive Screening Design to rank the relative importance of these factors and their interactions, then built and verified predictive models across solutions of 1 to 23 cp and target fill volumes of 0.2 to 2.0 mL.

Two implications follow for research peptides. First, low fill volumes sit at the difficult end of the range studied, and many research peptide vials are filled at or below 1 mL before lyophilization, where relative error is proportionally larger. Second, none of the significant factors are chemistry. A supplier with excellent synthesis and a rigorous purity programme can still produce inconsistent vial content if the fill operation is uncontrolled, because these are independent parts of the process.

What the Grey Market Evidence Actually Showed

The clearest published data on vial content in the unregulated peptide supply chain comes from Ashraf and colleagues, whose 2024 study in the Journal of Medical Internet Research combined market surveillance with laboratory analysis of test purchases. Their methodology is unusually complete for this space: search engine results page analysis, website content assessment, domain traffic analytics, visual packaging inspection, sterility and endotoxin evaluation, and quantitative analysis by liquid chromatography coupled with mass spectrometry.

Test purchases were attempted from six illegal online sellers. Three vials were delivered and three prefilled pen orders were never fulfilled at all. Every delivered vial was assessed as probable substandard or falsified product. The finding most relevant here is the content result: measured peptide content exceeded the labeled amount by 28.56 percent to 38.69 percent. The vials were overfilled relative to their stated content, and no peptide-related impurities were detected in them.

That last detail deserves attention, because it demonstrates the independence of these attributes precisely. The absence of peptide-related impurities indicates the peptide material itself was not degraded or heavily truncated. The purity figure of 7.7 to 14.37 percent therefore reflects overwhelming non-peptide mass, most plausibly excipient and residual counterion, which is the same failure pattern that quantitative NMR analysis is designed to expose and that chromatographic purity alone cannot detect. Content, purity, and identity failed in three different directions in the same vials.

Reading a Certificate With Sampling in Mind

Several practical checks follow from the standards and the published data. A certificate should identify the batch, the analytical laboratory, the date of analysis, and the methods used, because a purity figure without a validated analytical procedure behind it has no evidentiary weight. It should distinguish chromatographic purity from peptide content, since these answer different questions and are established by different methods. Where the certificate reports impurities, the reporting and identification thresholds matter, and those threshold conventions carry their own assumptions about detector response.

What no research peptide certificate currently offers is a unit uniformity result. That is an honest description of the state of the category rather than a criticism of any one supplier. Running a full uniformity assessment consumes 10 to 30 vials of saleable product per batch and is economically difficult at research supply volumes. The reasonable middle position is transparency about what was tested and what was not, so that a laboratory receiving material for a compound such as tirzepatide can determine peptide content in-house by amino acid analysis or quantitative NMR when the experimental design requires an absolute mass figure rather than a relative purity figure.

The general principle is straightforward. Treat the certificate as evidence about a batch, verified by a named third-party laboratory, and treat the individual vial as an untested member of that batch. Where an experiment depends on knowing the absolute quantity of peptide present, that quantity should be measured, not assumed from a label.

Sampling is one of several attributes a purity figure leaves undescribed. The handling of a result that fails its specification is covered in out-of-specification investigation and retesting limits, the physical integrity of the container itself in container closure integrity testing, and the fields worth checking on the document in how to read a certificate of analysis.

For research purposes only. Not for human consumption. Not for diagnostic or therapeutic use.