Peptide Purity: What the Numbers on Your COA Actually Mean

In 2024, 43% of peptides sent to third-party labs failed to meet the purity claims on their labels. Lower-tier vendors showed actual purities of 71–91% while advertising 99%+. But peptide purity is even more complicated than bad actors suggest: even when the number is accurate, most buyers misunderstand what it measures.

A 99% pure peptide can contain only 70–85% actual peptide by weight. A 100% pure peptide can have a net peptide content below 70%. The rest is water, counterion salts, and residual solvents that HPLC never sees. And purity tells you nothing about safety. In one published analysis, 65% of internet-purchased research peptides exceeded endotoxin safety thresholds, all while carrying purity certificates.

This piece breaks down how HPLC testing works, what each COA field means, how to calculate actual peptide content from a TFA salt vial, and how to spot fabricated certificates. By the end, you will know the six questions to ask any vendor before buying. Our vendor directory grades transparency so you don't have to guess.

What Peptide Purity Actually Measures

“Purity is about the peaks. NPC is about the powder.” That line from Tidelabs, a UK-based peptide QC lab, draws the sharpest distinction most buyers never learn.

HPLC purity is a ratio. The formula: (area of main peak / total area of all peaks) × 100. If the main peak has an area of 2,500,000 and the combined impurity peaks total 75,000, you get 2,500,000 / 2,575,000 = 97.1% purity. That percentage tells you what fraction of the UV-absorbing peptide material in the sample is your target compound. The detection wavelength, 214–220nm, is where peptide bonds absorb UV light. Anything that does not absorb at this wavelength is invisible to the measurement.

What HPLC Catches

The test detects synthesis-related impurities. Deletion sequences where an amino acid was skipped during coupling. Truncation sequences that stopped short. Oxidized variants, particularly on methionine and cysteine residues (+16 Da). Deamidated forms (asparagine to aspartate, glutamine to glutamate, +1 Da). Dimers. These are the byproducts of solid-phase peptide synthesis, and HPLC separates them by their interaction with the column.

What HPLC Misses Entirely

Water. TFA counterion salts. Acetate. Non-UV-absorbing contaminants. Endotoxins. Heavy metals. None of these register at the 214–220nm detection wavelength. As Dr. Hillary Lin, MD, puts it: “You can have a vial of 99.9% pure peptide that is contaminated with bacterial byproducts from a dirty lab.”

HPLC also cannot confirm identity. A peptide missing one amino acid may appear 99% pure but be 0% the desired compound. It cannot distinguish D- and L-amino acid isomers either. A peptide synthesized with the wrong chirality co-elutes with the correct version on standard columns. Only chiral HPLC or circular dichroism can catch that.

Peptide purity answers one question well: what fraction of peptide-related material is the target? It leaves safety, identity, and actual content unanswered. Our COA verification methodology evaluates whether vendors provide the measurements that fill those gaps.

How HPLC Testing Works: The Method Behind the Percentage

Every COA you've read cites HPLC. Here is what that machine actually does to your peptide sample.

The Setup

Reverse-phase HPLC pushes a peptide solution through a C18 silica column (3–5 μm particle size, 100–300 Å pore diameter, 150–250mm length) using a gradient of acetonitrile and water. The gradient typically runs from 5% to 95% acetonitrile over 30–60 minutes, with 0.1% TFA added as an ion-pairing reagent to sharpen peak shape. A UV detector at 214–220nm records signal as each compound exits the column.

Reading a Chromatogram

The x-axis shows retention time, meaning when a compound exits the column. The y-axis shows UV signal intensity. A high-purity peptide produces one dominant, sharp peak towering over everything else.

Multiple peaks of varying sizes mean impurities are present. Shoulder peaks deserve special attention. These are partially resolved impurities that merge with the main peak, and integration software counts them as part of the main peak area. That inflates the reported purity number without the analyst necessarily flagging it. Any peaks exceeding 1–2% of total area represent meaningful impurities worth investigating.

Why Method Matters for Comparison

Different columns, gradient lengths, mobile phases, and detection wavelengths produce different results. A peptide showing 99% on one method may show 97% on another with no actual quality difference. You cannot compare peptide purity numbers across different vendors using different labs. The numbers are only meaningful within the same analytical method.

The Shelf Life Caveat

An HPLC result applies only to the tested batch at the time of analysis. Over 12–24 months, degradation products form and can co-elute with the main peak on aged samples. The reported purity stays high while bioactivity quietly declines. Scrutinize any COA with an analysis date older than a year. A re-test is the only way to confirm current quality on stored peptides.

Purity vs. Net Peptide Content: The Number Most Buyers Overlook

Most dosing errors start here. A 10mg vial at 99% HPLC purity but 78% net peptide content contains roughly 7.8mg of actual peptide. If you make a 1 mg/mL solution using the full vial in 10mL solvent, your actual concentration is 0.78 mg/mL. That is a 22% error running through every data point in your experiment.

What NPC Measures

Net peptide content is the actual peptide mass as a percentage of total powder weight. The remainder is water absorbed during lyophilization, counterion salts (TFA or acetate), and residual solvents. Typical NPC range: 60–90% of gross weight. A 100% pure peptide can have a net peptide content below 70%.

The TFA Problem

Every basic amino acid residue (lysine, arginine, histidine) and the free N-terminal amine binds one TFA molecule during synthesis and HPLC purification. TFA has a molecular weight of 114 Da. These molecules add mass that HPLC never sees.

The theoretical NPC formula:

Take a peptide with MW of 1,000 and two basic residues in TFA salt form. NPC = 1,000 / (1,000 + (2 × 114)) = 1,000 / 1,228 = 81%. Now convert that same peptide to acetate salt (MW = 59 per counterion). NPC = 1,000 / (1,000 + (2 × 59)) = 1,000 / 1,118 = 89%. Same peptide. Eight percent more actual peptide per vial, just by changing the salt form.

A peptide with five lysine residues and one arginine carries six TFA counterions. At 114 Da each, that is 684 Da of TFA riding along. For a 3,000 Da peptide, gross weight becomes 3,684 Da, and NPC drops to 81.4%.

Why It Matters for Buyers

Standard labels refer to gross weight. A 5mg vial at 81% NPC contains 4.05mg of actual peptide. Use NPC for concentration calculations, not the label weight. If the COA does not report NPC, use the theoretical formula above as an estimate, then count your basic residues.

Acetate vs. TFA Salt Forms

Most FDA-approved peptide drugs use acetate salts. The reason is partly safety (TFA can inhibit cell growth at nanomolar concentrations in some cell lines) and partly content: acetate salts deliver more peptide per milligram. TFA removal costs 20–30% more due to peptide loss during conversion, but for cell-based assays, the cleaner salt form can prevent confounded results.

How to Read a Peptide COA: Field by Field

Every certificate of analysis contains the same core fields. Here is what each one tells you, and what to flag when a field is missing.

Product ID and Sequence

Confirm the peptide name and amino acid sequence match your order.

Batch/Lot Number

This must match the number on your vial. If it does not, that COA does not verify your specific product. Some vendors recycle a single COA across multiple batches. A mismatched lot number is the single fastest way to identify a useless certificate of analysis.

Analysis Date

Scrutinize anything older than 12 months. Lyophilized peptides degrade, and a stale COA may not reflect your batch's current state.

Testing Laboratory

Look for a named lab with verifiable credentials. Third-party testing is preferred over in-house. One distinction matters: a vendor naming a lab on their COA does not mean that lab is accredited. ISO/IEC 17025 is the international standard for testing lab competence, requiring externally audited procedures and calibrated equipment. Our methodology v2.0 scores this difference. A named lab is not the same as an accredited lab.

HPLC Purity Percentage

Look for 98% minimum. Confirm the COA includes a chromatogram image, not just a number. Without the graph, you cannot check for shoulder peaks or baseline issues that inflate reported purity.

Mass Spectrometry

The observed molecular weight should match the theoretical value within ±0.5 Da. A note on notation: [M+H]+ is a singly charged ion. [M+2H]2+ is doubly charged and appears at half the actual mass on the spectrum. This is normal for larger peptides and does not mean the peptide is half the expected size.

Net Peptide Content

If provided, use this number for dosing calculations instead of gross weight. If missing, use the theoretical formula from the previous section.

Endotoxin

Critical for cell-based work. Acceptable limits: less than 1 EU/mg for general cell culture, less than 0.1 EU/mg for immunological research involving macrophage activation, cytokine assays, or NF-κB pathway studies. Most research-use-only vendor COAs do not include endotoxin data. That absence is itself a data point.

Six Questions to Ask Any Vendor

Are HPLC and mass spectrometry both present? Is the lab third-party? Does the batch number match your vial? Is endotoxin data available? Is NPC reported for this lot? What is the storage and cold chain protocol? A “no” on any of the first three should stop a purchase. See how vendors compare on these criteria in our vendor directory.

98% vs. 99% Purity: When the Difference Actually Matters

Short answer: the difference is application-dependent, not universal. Chasing 99.5% when you need 98% wastes money. Using 95% when you need 99% wastes data.

Thresholds by Application

When 1% Actually Changes Your Results

At sub-nanomolar concentrations, a 1% impurity fraction represents a proportionally significant share of total material. In competitive binding assays, truncation sequences with partial binding affinity can produce falsely lower apparent affinities for the target compound.

The hemopressin case illustrates the extreme scenario. Hemopressin is a CB1 receptor antagonist. Related impurity peptides found in the same synthesis batch turned out to be CB1 receptor agonists. Opposite activity from the same impurity pool. At 1–2% impurity levels, which would appear as a 98–99% pure sample on HPLC, the functional readout of any CB1 cell-based assay would be confounded. Published in Drug Discovery Today (PMID 25044089).

Consistency Beats Marginal Purity

Finnrick Analytics tracks 5,930 samples from 196 vendors. Their data reveals something more important than average peptide purity: batch-to-batch consistency. Top vendors like Paradigm Peptide (8.9 average across 16 products) deliver reliable results. HK Peptides scored anywhere from 0.0 to 10.0 across 342 tests, averaging 7.0. Deuschem averaged 4.1 across 29 tests. A vendor that consistently delivers 98% is more reliable than one averaging 99% with wild variance. The COA on last month's order tells you nothing about this month's batch.

How to Spot a Fake or Misleading COA

A vendor called Modern Peptides posted a NAD+ COA with the Janoshik verification key and QR code deliberately removed. When the test number was looked up, the original test corresponded to “QSC BAC water.” Bacteriostatic water. Not a peptide at all. Janoshik publicly called out the vendor by name. The fraud was independently confirmed by TrustPointe.

That case was caught. Most are not. As one GLP-1 Forum user observed: “The majority of their customers probably WON'T know, and that's what matters.”

Red Flags That Should Stop a Purchase

• No digital verification. No task ID, QR code, or portal link. You have no way to confirm the document is authentic.
• Lot number mismatch. The number on the vial does not match the COA. That certificate does not cover your product.
• Purity reads 100.0%. Analytically impossible. Even 99.99% warrants skepticism.
• Identical numbers across products. Every peptide, every batch, every month shows the same purity. That is a template, not a test result.
• HPLC number but no chromatogram. A purity claim without supporting data.
• No mass spectrometry. Without MS, you have a purity number for an unidentified compound.
• Testing lab absent or unverifiable. If you cannot find the lab online, the COA carries no weight.
• COA not tied to a specific batch. Generic certificates that lack lot numbers apply to nothing.

Verifying a Janoshik COA

Find the Task Number at the top-left of the report. Find the Unique Key at the bottom (case-sensitive, all capitals). Enter both at public.janoshik.com. The system confirms whether the report matches its database. If the verification returns a result for a different product, or if the key was stripped from the document, walk away.

What Third-Party Testing Labs Can and Cannot Verify

That 43% failure rate from 2024? Third-party testing caught what vendor COAs hid. Independent labs are the reason we know the gap exists between claimed and actual peptide purity.

What Independent Labs Test

A full panel from a lab like Janoshik includes HPLC purity, LC-MS/MS identity confirmation, LAL endotoxin testing, ICP-MS heavy metal screening, and TAMC/TYMC sterility testing. A GLP-1 peptide HPLC test runs about $300. The full panel costs roughly $828, plus $50–75 for US shipping.

Accredited vs. Named Labs

ISO/IEC 17025 is the international standard for testing lab competence. It requires externally audited procedures, calibrated equipment, and documented methods. A vendor naming a lab on their COA does not mean that lab holds ISO/IEC 17025 accreditation. This distinction matters. Accredited results carry weight in regulatory contexts. Named-but-unaccredited results are better than in-house testing, but they are not equivalent to pharmaceutical-grade QC.

Janoshik: Industry Standard With Limits

Janoshik is the gold standard for gray market verification. The unique key verification system, decade-plus track record, and full-panel availability make it the most recognized name in independent peptide testing. But it operates without ISO 17025 accreditation. COAs lack methodology details. Raw chromatographic data sits behind a paywall. Results are not suitable for regulatory submissions. And a February 2026 data breach exposed customer information, raising operational security questions for anyone sending samples with personal details.

No Lab Is Infallible

Chromate Lab misidentified multiple peptide samples in 2025. Semaglutide samples were labeled as containing capromorelin. Retatrutide samples were reported as tirzepatide. Other labs, including Janoshik, Freedom Diagnostics, and TrustPointe, correctly identified those same samples. Lab reliability varies, and no single result from any single lab should be treated as gospel.

The Bottom Line on Third-Party Testing

Independent testing eliminates the vendor's financial incentive to inflate data. That is its primary value. Our vendor directory tracks which vendors provide third-party COAs with verifiable lab credentials.

Beyond Purity: The Tests a Complete COA Should Include

In a 2023 analysis of 84 RUO GHK-Cu products, 37% failed basic identity confirmation via mass spectrometry. Six products contained less than 5% actual GHK-Cu. The rest was filler, primarily mannitol and glycine. A filler product could theoretically show “high purity” on HPLC if that filler elutes as a single clean peak.

Peptide purity is one measurement. Here is the full picture.

Mass Spectrometry for Identity

Confirms the correct molecular weight within ±0.5 Da. Without it, HPLC purity is meaningless. You might have 99% pure wrong compound. A study in Drug Testing and Analysis found 30% of internet-purchased research peptides contained incorrect amino acid sequences. HPLC alone would never have flagged them.

Endotoxin Testing

In one published analysis, 65% of internet-purchased research peptides exceeded safety thresholds for endotoxin contamination. Endotoxins trigger inflammation at 0.1 ng/mL in cell culture. Most RUO vendor COAs do not report endotoxin levels. If you are running cell-based assays, this missing data point matters more than whether purity is 98.5% or 99.1%.

The Complete Quality Picture

Purity (HPLC) + Identity (MS) + Content (NPC) + Safety (endotoxin, sterility, heavy metals) = the full assessment. Most vendor COAs provide only one or two of these. The number of tests a vendor routinely includes on their certificates is a transparency signal. Vendors offering only HPLC purity are giving you the minimum. Vendors providing four or more measurements are telling you something about their standards.

This is why we grade vendor transparency, not peptide quality. We do not test peptides. We evaluate whether vendors are willing to show you the receipts. See how vendors compare on our directory.

FAQ

What does 99% peptide purity actually mean?

99% of UV-absorbing peptide material detected by HPLC is the target compound. The other 1% is synthesis impurities. This does not mean 99% of the vial weight is peptide. Water, TFA salt, and acetate counterions are invisible to HPLC.

Is 98% purity good enough for research?

For most cell-based assays, yes. 98% is the accepted minimum for receptor binding, cAMP, cell viability, and migration studies. 99%+ is needed for mechanistic studies and publication-grade dose-response curves.

What is the difference between peptide purity and net peptide content?

Purity measures target peptide versus other peptide impurities. NPC measures actual peptide mass versus everything in the powder (water, salts, counterions). A 99% pure peptide typically has 60–85% NPC. Both numbers are needed for accurate dosing.

Does high HPLC purity mean a peptide is safe?

No. HPLC does not detect endotoxins, heavy metals, or sterility issues. 65% of internet-purchased research peptides exceeded endotoxin thresholds despite carrying purity certificates.

How do I verify if a COA is real?

For Janoshik COAs, enter the Task Number (top-left) and Unique Key (bottom, case-sensitive) at public.janoshik.com. If the QR code or key is missing, the vendor likely removed it to prevent verification. Cross-reference lot numbers on vial and COA.

Why does my 5mg vial contain less than 5mg of peptide?

The label shows gross weight including water, counterion salts, and residual solvents. Net peptide content runs 60–90% of gross weight. For TFA salt peptides with many basic residues, TFA alone reduces actual content by 5–25%.

What is TFA and why does it affect peptide weight?

Trifluoroacetic acid (MW = 114) binds to every basic residue during synthesis and purification, adding mass invisible to HPLC. Converting to acetate salt (MW = 59) yields higher net peptide content. Most FDA-approved peptide drugs use acetate.

How common is it for vendor purity claims to be inaccurate?

In 2024, 43% of peptides tested by third-party labs failed label claims. Lower-tier vendors showed 71–91% actual purity while claiming 99%+. Finnrick data across 5,930 samples reveals wide batch-to-batch variance. A single good COA does not guarantee the next batch.