What '98% Purity' Actually Means on a Peptide COA (2026)

"98% purity." It is printed on more research-peptide labels and Certificates of Analysis than almost any other figure, and it reads like a hard fact — as if an instrument inspected the vial and announced it was 98% peptide. That is not what the number says. A purity percentage is a specific, narrow statement built on a basis and a set of assumptions, and reading it well means knowing exactly what it does and does not claim.

This is research-use educational content. Nothing here is a dosing recommendation or human-use claim — the focus is on interpreting a number, not acting on it.

What the number is measured on

In nearly every research-peptide COA, a purity figure is an HPLC area-percent result. The instrument separates the sample's components in time, a detector records each one as a peak, and software integrates the area under every peak. Purity is then the target peak's area divided by the total area of all detected peaks. A 98% result means the main peak accounts for 98% of the combined detected area — no more, no less. For the mechanics of how that calculation is made, our peak integration walkthrough goes line by line; for the method itself, see what HPLC is.

The word basis matters because area-percent is a relative measure. It compares peaks to other peaks. It is not a count of milligrams, and it quietly assumes every component absorbs UV light at a similar rate — a fair approximation for peptides and their close relatives, but an assumption nonetheless.

What the other 2% is — and what it is not

The non-target fraction is real material, and on a credible chromatogram you can usually see it as small peaks flanking the main one. It typically includes truncated or deletion sequences left behind by imperfect synthesis, oxidation and degradation products that form during processing or storage, and occasionally co-eluting impurities the method could not fully resolve from the main peak. None of that is alarming at the 2% level; no real synthesis is perfectly clean.

What the 2% is not is just as important. Area-percent only counts what the detector registered. Counterions, residual water, and salts that ride along with lyophilized peptide are largely invisible to this measurement — they are not the 2%. That non-peptide weight is a different question entirely, belonging to potency or net peptide content rather than purity. A vial can read 98% pure and still hold less actual peptide than its label, because purity never measured the absolute amount. We separate those two ideas fully in purity vs potency.

Why 98% is a normal result, not a red flag

There is a common instinct to treat anything below 99% as a near-miss. For synthesized research peptides, that instinct is miscalibrated. Solid-phase peptide synthesis accumulates small imperfections at every coupling step, so real, honestly reported HPLC figures land in a band — roughly 95% to 99% for most compounds. A 98% result sits squarely inside that range. It is the kind of number a working, transparent lab produces routinely.

The figure that should raise an eyebrow is the too-clean one: an identical 99.9% printed across an entire catalog, or a flat 100%, neither of which is consistent with how synthesis and analysis actually behave. As we cover in why most peptide COAs are worthless, implausibly perfect numbers with no chromatogram behind them are a stronger warning sign than a believable 98% with a visible trace.

98 versus 99: a smaller gap than it looks

So is 98% meaningfully worse than 99%? On paper, one point. In practice, often less. The exact number depends on how the chromatogram was integrated — where the baseline was drawn, how peak start and end points were set, whether two overlapping peaks were split or merged. Two competent analysts can integrate the same raw data and report figures a fraction of a point apart. A single percentage point of headline purity can sit inside that integration noise.

That is why the honest comparison is rarely "98 versus 99." It is "supported versus unsupported." A transparent 98% accompanied by a batch-specific chromatogram, a stated method, and a retention time tells you far more than a bare 99.9% with nothing behind it. The supporting data is the signal; the last decimal place mostly is not.

How to read a 98% figure well

A few practical habits turn the headline number from a slogan back into information:

• Confirm the basis. It should be stated as HPLC area-percent (or similar). A purity number with no named method is a claim, not a result.
• Look for the chromatogram. The peak data is the source; the percentage is just a calculation from it. No trace means the number is unverifiable — see the visual chromatogram guide.
• Check that purity is not standing in for everything else. Purity says nothing about identity (is it the right molecule?), potency (how much is there?), or sterility (is it biologically clean?). A complete picture needs the three separate tests, not one figure carrying all the weight.
• Match the batch. A 98% that belongs to a different lot than the one in your hand tells you about a vial, not your vial. Our how to read a COA guide covers the full document.

For independent confirmation, sending a sample to an outside lab is the cleanest check — our research methods hub and the peptide catalog frame how to evaluate a source on its full data package rather than a single percentage.

Bottom line

A 98% purity figure on a peptide COA is an HPLC area-percent result: the target peak is 98% of the combined detected peak area, and the other 2% is other detected components — not the salts and water that affect potency. For a synthesized research peptide, high-90s purity is a normal, credible band, and a believable 98% backed by a visible chromatogram beats an unsupported 99.9% every time.

Read the number for what it is — a relative statement about peaks, on a stated basis, for a specific batch — and never let it stand in for identity, potency, or sterility. Pair this with LC-MS for identity confirmation and reference standards in HPLC to read a COA on the whole picture. For research use only.