HPLC vs LC-MS: which peptide test answers what

Certificates for peptides often list "HPLC" and "LC-MS" side by side, and some sellers quote only one. The two share a front end, liquid chromatography, which pushes the dissolved sample through a column so its components come out one after another. What differs is the detector at the end, and the detector decides which question the test can answer.

What question does each test answer?

The short version: HPLC is the ruler for how much and how clean, and LC-MS is the scale that says what. The lessons on HPLC purity and identity go deeper into each one. This page is about choosing between them and reading them together.

Why can HPLC alone not confirm identity?

A UV detector sees anything that absorbs ultraviolet light at the chosen wavelength, and most peptides do. It reports when each peak arrives and how big it is. It does not report what the peak is. If a vial labelled as one peptide actually holds a different peptide that is just as clean, the UV trace can show a single tall peak and a high purity figure.

Retention time, when the peak arrives, can support identity if the laboratory ran a reference standard of the labelled peptide under the same conditions. That is why the European Medicines Agency lists relative retention time among acceptable identification methods. It also recommends that identification of a synthetic peptide use "at least two orthogonal methods", naming mass, relative retention time, LC-MS, peptide mapping, amino acid analysis and NMR as options (EMA, synthetic peptides guideline, 2025). Orthogonal means the methods rest on different measurement principles, so one can catch what the other misses.

What does LC-MS add that HPLC cannot?

A mass number. The mass spectrometer turns each component into charged ions and measures their mass-to-charge ratio. For a peptide of known sequence, the expected mass can be calculated, so the measured value either matches it within the instrument's tolerance or it does not.

A published reference-standard study shows how close that match can be. For leuprolide, the theoretical mass-to-charge value of the protonated molecule was 1209.6533 and the measured value was 1209.6515, and fragmentation data covered the full amino acid sequence (McCarthy et al., Pharmaceutical Research, 2023). The international validation guideline notes that for some technologies, specificity can be predicted from technical parameters, giving "resolution of isotopes in mass spectrometry" as an example (ICH Q2(R2)).

Can LC-MS find impurities that HPLC misses?

Yes, and this is the main reason to want both. Peptide impurities are often the target peptide with one amino acid missing, added or changed. They behave almost the same in the column, so they can come out at nearly the same time as the main peak and be counted as part of it.

A study of synthetic glucagon describes exactly this. The authors found that peptide impurities with sequences similar to glucagon had retention times close to the main component, and that small impurity peaks were "frequently masked by the main component peak". They used LC-MS to tell the impurities apart by mass instead (Wang et al., Scientific Reports, 2020). The EMA guideline makes the same point for regulators, asking that methods for identity and impurities account for "the risk of co-eluting impurities".

Why is HPLC still the usual source of the purity number?

Because a UV detector responds to peptide bonds in a broadly similar way across related peptides, so peak areas can be compared directly as percentages. A mass spectrometer's response depends on how easily each molecule forms ions, which varies from one molecule to the next. In the glucagon study, quantifying each impurity by mass spectrometry needed a separate synthetic standard and calibration curve for every one of the nine impurities measured.

For amount, the same logic applies. The EMA guideline describes assay as usually based on "a chromatographic method usually the same as for purity" run against a reference standard. The reference-standard study above used the compendial HPLC assay to assign milligrams per vial. LC-MS can quantify, but a routine amount figure on a peptide certificate normally comes from HPLC against a standard.

When do you need both?

• When substitution is the worry. High-value compounds such as retatrutide are where swaps are most tempting. Purity cannot show a swap; mass can.
• When you want to compare vials by milligram. Identity tells you it is the right molecule; HPLC against a standard tells you how much of it there is.
• When a purity figure looks too good. A very high purity with no identity data says only that one peak dominates.
• When a certificate will be used in a dispute. A report with two orthogonal methods is harder to argue with than one with a single figure.

The international guideline puts it generally: "Where one analytical procedure does not provide sufficient discrimination, a combination of two or more procedures is recommended" (ICH Q2(R2)).

How do you tell which one a report used?

Look for the method next to each result, not only in a footer. An identity line should show an expected mass and an observed mass, or a statement that the spectrum matched. A purity line should name HPLC or UPLC and, ideally, the detection wavelength. If a report gives "purity: 99%" and "identity: confirmed" with no method or masses, you cannot tell which instrument produced either line. The COA checker walks through these fields, and the which test tool helps you decide which methods to order.