Why lab tests disagree: method, sample and timing

A lot is tested by one lab at 99.3%. Someone sends a vial to another lab and gets 97.8%. One of them must be wrong, surely?

Not necessarily. "Pure by what method?" is the first question to ask, and there are two more after it: which sample, and when. Most disagreements trace back to one of those three.

1. Method

The same material can read differently under different, perfectly valid methods.

Column and gradient. A shallower gradient or a different column chemistry can separate an impurity that another method hides under the main peak. Better separation usually means a lower purity figure, not worse material.

Detection wavelength. At 214 nm, the peptide bond absorbs strongly and most peptide impurities are visible. At 280 nm, only aromatic residues absorb, and impurities without them disappear. A figure at one wavelength cannot be compared with one at another.

Integration. Where the baseline is drawn, how shoulders are split, and the smallest peak counted all change the denominator. Two analysts on the same data can differ by a few tenths of a percent.

Run length. A short run can end before late impurities come off.

What is being measured at all. One report gives HPLC area percent, another gives net peptide content, a third gives an assay against a standard. These are different quantities that happen to share a percent sign.

2. Sample

Different vials. Two vials from one lot are usually close, but not identical. Fill, sealing and handling vary.

Different lots. Surprisingly common in comparisons. Check that both reports carry the same lot code.

Sample preparation. The diluent, concentration and how long the sample sat before the run can all shift peaks or create small artifacts.

Amount of material. Tests that need very small amounts are more sensitive to weighing and handling error.

3. Timing

Age of the material. A lot tested at release and the same lot tested months later may genuinely differ, especially for oxidation-prone sequences or material that saw heat or moisture.

Conditions in between. Shipping, storage temperature and how often a vial was opened all matter.

Age of the method. Labs update methods. A report from last year may not use the same method as one from today, even at the same lab.

Comparing two reports fairly

Check	Why
Same lot code?	Otherwise they describe different material
Same quantity?	Area percent, content and assay are not comparable
Same wavelength and similar method?	Wavelength alone can move the figure a lot
How far apart in time, and how was the material held?	Real change is possible
Full chromatograms available?	The chart often explains the difference at a glance

When a difference is meaningful

Small differences under different methods are expected. A large gap, a new impurity peak that matches a known degradation product, or a mismatch in identity (mass) is worth following up with both labs and the supplier. Identity disagreements in particular are not explained by method choice.

None of this is a reason to distrust testing. It is a reason to read every figure alongside its method, its sample and its date.