Purity 98 or 99 percent: what the number on a peptide certificate does and does not tell you

Written 23 September 2026, last changed 7 October 2026 by the trade desk. About 4 minutes to read.

A purity of 98 or 99 percent is the main peak’s share of the HPLC peak area at 220 nm, not the share of the vial’s weight: a 10 mg vial at 99 percent still holds 10 mg of material, water and acetate included. The release limit for a peptide line here is 98.0 percent.

What is being measured

The sample is run through a reverse-phase HPLC column and the peaks are recorded at 220 nm. Purity is the area of the main peak as a share of the area of all the peaks. Our release limit is not less than 98.0 percent, and any single related substance not more than 1.0 percent. That limit is for the peptide lines; protein, glycoprotein and preparation lines are released on the methods their own pages name, as the quality page sets out.

What the rest is

Related substances: sequences missing a residue, sequences with an extra one, oxidised or deamidated forms. They are structurally close to the target, a residue or a few atoms away from it. A 99 percent result means those together are one percent of the peak area.

What the number is not

It is not the weight of peptide in the vial. The dry cake also contains water, up to 8 percent by our limit, and the acetate counter-ion. If you need to know how much peptide by weight is in a vial, that is net peptide content: a separate figure, which the desk answers for a particular batch.

Why identity comes first

A pure peak of the wrong molecule is still the wrong molecule. Identity is confirmed by mass spectrometry, which checks the measured mass against the calculated one. Purity without identity is a number without a subject. What each result is, and what telling two close lines apart takes: HPLC and mass spectrometry on a peptide certificate.

98 against 99 in practice

Both are inside our limit. The difference is one percentage point of related substances, which is a smaller difference than whether the identity and net content results pass. Treat a certificate as a whole rather than as one number.

Below 98 percent: what happens to the batch

A batch whose purity comes in under 98.0 percent goes into quarantine. It is not blended with a better batch, not sold cheaper, and not offered under another label.

If your own laboratory tests a vial and reads lower than the certificate, send the desk the batch number and your result. We pull the retained sample for that batch, held twelve months from the batch's last dispatch, and run it against the same specification. Out of specification, the batch is withdrawn and every account that received it is told. In specification, you get the data.

Before comparing two numbers, compare the two methods. Area percent moves with the detection wavelength, the gradient, the column, the integration threshold and the reference standard. Two laboratories running different methods report different figures for the same vial, and neither is wrong; the specification names its method.

Applies to

This guide was written with Semaglutide and BPC-157 in mind, and it holds for the other lines of their kind; the full list is on the compounds page.

Questions about the purity figure

What is the other two percent in a 98 percent peptide?

Related substances, and they are relatives of the target rather than dirt from somewhere else: sequences missing a residue, sequences carrying an extra one, oxidised forms and deamidated forms. At 98.0 percent those together are two percent of the peak area, and the specification caps any single one of them at 1.0 percent as well, so a batch cannot meet the headline figure by carrying one large impurity.

Which matters more on a certificate, purity or identity?

Identity, because a pure peak of the wrong molecule is still the wrong molecule. On our certificates the measured mass is matched to the calculated average mass, within about one unit. Purity without identity is a number without a subject, so the identity line is the one to read first.

What happens to a batch that tests below 98 percent?

It goes into quarantine. It is not blended with a better batch, not sold cheaper and not offered under another label. If your own laboratory reads lower than the certificate, send the desk the batch number and your result: the retained sample for that batch, held twelve months from its last dispatch, is pulled and run against the same specification, and out of specification the batch is withdrawn.

Why do two laboratories report different purity figures for the same vial?

Because an area percentage moves with the method. The detection wavelength, the gradient, the column, the integration threshold and the reference standard all change the figure, so two laboratories running different methods report different numbers for one vial and neither of them is wrong. That is why a specification names its method, and why two figures are worth comparing only once the two methods have been.