No published laboratory test we found can prove that a jar of ghee is A2. Whether milk is A1 or A2 depends on a protein, beta-casein, set by the cow's genes, and the validated tests read the cow's DNA or the protein in its milk. Ghee is almost pure fat, so the evidence for an A2 claim has to come from the herd and the milk, before the ghee is made.
That is not a reason to distrust every A2 label. It is a reason to know what kind of evidence can stand behind one, and which tests on a ghee report answer which questions.
At a glance
- A1 and A2 are two forms of beta-casein, a milk protein, set by the cow's genes; every validated test reads the cow's DNA or the protein in milk.
- Methods published for milk include PCR genotyping, PCR on raw milk (detecting A1 down to 2 per cent), isoelectric focusing, LC-MS/MS (5 per cent detection limit) and ELISA; none has been validated on ghee.
- Ghee must be at least 99.5 per cent milk fat under FSSAI's standard, leaving at most 0.5 per cent for protein, water and everything else.
- FSSAI's ghee standard has no A1 or A2 parameter, so a ghee lab report cannot show A2 status; ours does not claim to.
- The useful evidence for an A2 claim is herd genotyping (every cow A2A2), milk screening for A1, and a batch report that meets FSSAI's limits.
Can a lab test tell whether ghee is A2?
Not reliably, on the evidence we found. Every published method we found for telling A1 from A2 was developed and checked on cows or on milk, not on ghee. Ghee must be at least 99.5 per cent milk fat under FSSAI's standard, and the A1 and A2 difference lies in a protein, not in the fat.
Search for "A2 ghee test" and you will find home tests, purity tests and lab reports offered as proof. None of those measure beta-casein. The tests that do measure it, described below, are research and quality-control methods for milk. Some can find small amounts of A1 in milk sold as A2. That is useful for a dairy checking its milk, and it is where the honest answer for ghee starts.
What makes milk A1 or A2?
One building block in one protein. Beta-casein is one of the main proteins in cow's milk, and its two common forms differ at position 67 of the protein chain: histidine in A1, proline in A2. Which form a cow makes is inherited, so A2 milk means milk from cows that carry two copies of the A2 form of the gene.
Because the variant is inherited, the evidence for it lies with the cows. A herd that is mostly A2 does not give A2 milk: one Polish study found more A1 than A2 protein in the milk and milk products it tested, even though the local Holstein cattle carried more A2 copies of the gene (63 per cent) than A1 (37 per cent). Milk is A2 only if every cow supplying it is A2A2.
Whether the difference matters for health is a separate question, and a weaker one than the marketing suggests. Our article on what A2 milk is and what the trials found sets out that evidence. This article is only about testing.
How do labs test milk for A1 and A2?
There are two broad routes. One reads the gene, using PCR on DNA from the cow or from cells in raw milk. The other reads the protein in milk, by isoelectric focusing, liquid chromatography with mass spectrometry, or antibody tests such as ELISA. Each has been described in peer-reviewed papers, all of them on milk.
The methods we found fall into these groups:
- PCR genotyping of the cow. An Austrian university laboratory found allele-specific PCR reliable for telling apart the common beta-casein variants (A1, A2, B and C), and used two further PCR methods to pick up variant A3 and to separate A2 animals from the rest (Mayer et al., 2021).
- PCR directly on raw milk. A Japanese team amplified the beta-casein gene straight from raw milk without extracting DNA first. Its results for 27 milk samples matched genotypes from the cows' DNA, and it could detect A1 down to 2 per cent in A2 milk (Watanabe et al., 2023).
- Isoelectric focusing. The Austrian study also used isoelectric focusing of milk proteins, which separates the protein variants themselves rather than the genes.
- Mass spectrometry. A Chinese team measured peptides specific to A1 and A2 beta-casein in cow milk by HPLC-MS/MS, and could estimate the proportion of each with a detection limit of 5 per cent (Wang et al., 2025). An earlier method quantified both forms in pasteurised milk (Guo et al., 2022).
- Antibody tests. A Polish study used ELISA to measure A1 and A2 beta-casein in raw, pasteurised and UHT milk and milk powder (Kamiński and Cieślińska, 2025). A Brazilian team produced an antibody that recognised A1 beta-casein in milk with 95.2 per cent sensitivity; some of its authors are partners in a biotech company (de Jesus et al., 2023).
Side by side, this is what each test reads and where it has been shown to work:
| Test | What it detects | Works on milk? | Works on ghee? |
|---|---|---|---|
| PCR genotyping of the cow | The cow's beta-casein genes: A1A1, A1A2 or A2A2 | Tests the animal, which is what makes its milk A2 | Not applicable; it tests the cow, not the jar |
| PCR on raw milk | A1 gene in cells in the milk, down to 2 per cent in one study | Yes, raw milk | Not tested in any study we found |
| Isoelectric focusing | Beta-casein protein variants | Yes | Not tested in any study we found |
| Mass spectrometry (LC-MS/MS) | Peptides specific to A1 and A2 beta-casein | Yes, cow milk, including pasteurised milk | Not tested in any study we found |
| ELISA and other antibody tests | A1 and A2 beta-casein protein | Yes, raw, pasteurised and UHT milk and milk powder | Not tested in any study we found |
| FSSAI ghee parameters | Moisture, milk fat, free fatty acids, identity of the fat | No; set for ghee | Yes, but says nothing about A1 or A2 |
The milk tests are not idle. The Austrian team found traces of A1 beta-casein in commercial "A2 milk" sold in Austria, and called for checks on farms and in shops.
Why is ghee so hard to test for A1 and A2?
Because the thing being tested is mostly gone. FSSAI requires ghee to be at least 99.5 per cent milk fat, made by processes that remove almost all the water and milk solids-not-fat. That leaves at most 0.5 per cent for moisture, protein and everything else together, and beta-casein is only one of the proteins.
The protein methods in the table were developed for milk, where beta-casein is plentiful. The products tested ranged from raw milk to UHT milk and milk powder. The milk PCR method in the Japanese study used raw milk. We found no paper that applied any of these methods to ghee, butter or another milk fat and showed that the result was reliable. Our article on how much protein ghee has explains where the milk protein goes when ghee is made.
There is a further gap even in milk. The mass spectrometry method above reports a 5 per cent detection limit for each form, and the raw-milk PCR method 2 per cent for A1. A result below those limits means "not detected", not "absent". If you are offered an "A2 ghee test", ask two questions:
- Which method does it use: DNA, protein, or something else?
- Has that method been validated on ghee, with a published detection limit?
If the answer to the second question is no, the result tells you little about the jar in front of you.
What does a ghee lab report actually check?
Quality and identity, not protein type. FSSAI's ghee standard sets limits for moisture, milk fat, free fatty acids, the butyro-refractometer reading, the Reichert-Meissl and Polenske values, iodine and saponification values, the Baudouin test, beta-sitosterol and fatty acid composition. None of these concerns beta-casein, and the standard has no A1 or A2 parameter.
Those tests answer real questions: is the ghee dry, is it fresh, and is the fat milk fat rather than vegetable fat? They cannot say which cows the milk came from. Our guide on how to read a ghee lab report takes each value in turn.
We hold our own reports to the same reading. Himdhara's batch reports are tests of the ghee, so we do not present them as proof of A2 status.
What can a buyer reasonably check instead?
Ask about the evidence behind the claim at the stage where it can be tested: the cows. The Austrian researchers recommend genotyping every cow at farm level to confirm A2A2 status, and screening products on the market for A1. For ghee, you can then check the batch report for the FSSAI parameters.
In practice, that means a short list of questions:
- Herd genotyping. Has every cow supplying the milk been genotyped as A2A2, and by which laboratory?
- New animals. How are new cows, and bulls used for breeding, checked before their milk is added?
- Milk testing. Is the milk itself ever screened for A1, by PCR or a protein test?
- Breed. Which breed is it? Breed is a strong hint about A2 status but not a guarantee, as our A2 milk article explains.
- The batch report. Does it carry a batch code that matches your jar, and does it meet the FSSAI limits?
A seller who answers the first three plainly is giving you more than any single "A2 ghee test" could. A seller who offers only a certificate for the ghee itself has given you a document whose method you should question.
The short version
A1 and A2 describe two forms of beta-casein, a milk protein, and which form a cow makes is set by her genes. The published tests read the cow's DNA or the protein in milk, by PCR, isoelectric focusing, mass spectrometry or antibody tests, and we found none validated on ghee. Ghee must be at least 99.5 per cent milk fat, and FSSAI's ghee standard has no A1 or A2 parameter. The useful evidence for an A2 ghee claim is herd genotyping, backed by a batch report that checks the ghee against FSSAI's limits.
Sources
Further reading behind the figures above. Himdhara's own numbers — yields, batch results — come from our production records and the batch reports on the lab reports page.
- Mayer HK, Lenz K, Halbauer EM (2021). "A2 milk" authentication using isoelectric focusing and different PCR techniques. Food Research International.
- Watanabe A, Munakata K, Muto M, Kuramoto T (2023). A highly sensitive PCR method for A1 allele detection in A2 milk samples without DNA isolation. Animal Science Journal.
- Wang F et al. (2025). Determination of A1 and A2 β-casein in cow milk by high-performance liquid chromatography tandem mass spectrometry. Journal of Dairy Science.
- Kamiński S, Cieślińska A (2025). β-Casein A1 and A2 Genetic Variants and β-Casomorphin-7 in Raw Milk and Processed Milk Products. International Journal of Molecular Sciences.
- Food Safety and Standards Authority of India. Food Safety and Standards (Food Products Standards and Food Additives) Regulations, 2011: Chapter 2.1, Dairy products and analogues (2.1.8 Standard for Milk Fat Products). FSSAI.