Same blood. Different number.

There isn't one universal AMH test. The same tube of blood can come back with different numbers depending on which lab machine measured it — Roche, Beckman, or an older manual ELISA each read a bit differently.

Compared head-to-head, the Roche Elecsys machine tends to report a lower AMH number than the older Gen II ELISA on the very same sample — the gap between the two methods has been measured anywhere from about -25% to +45%.

The Beckman Access test often reads higher than the Roche Elecsys test on the same blood — different enough that in one study around 29% of women would have been sorted into a different medication-dosing band just because of which machine was used.

When researchers ran the exact same samples on both the Roche and Beckman automated machines, the numbers still differed by a few percent and in a consistent direction, even though the two tests tracked each other closely.

It's not only the brand of machine — the same test run in different labs can disagree a lot. In one quality-check program, individual results scattered from about 38% below to 55% above the true value, so it's best to stick with one lab over time.

For years there was no universal 'gold ruler' for AMH, which is exactly why the machines disagree. The WHO finally issued a shared reference material in 2020, but it didn't work equally well on every machine — so full standardization is still a work in progress.