Gut microbiome tests promise a picture of your gut bacteria, a health score and a personalized diet. They cost from about 100 to 500 US dollars, and they do not all measure the same thing. In 2026, a study by the US National Institute of Standards and Technology (NIST) also showed that they can disagree on the very same stool sample. Here is how to choose, if you decide to test at all.
Three different methods
| Method | What it reads | What it can tell you | Examples |
|---|---|---|---|
| 16S rRNA sequencing | One marker gene shared by all bacteria | Which groups of bacteria are present, mostly at genus level. Cheapest. | Ombre (formerly Thryve), Biomesight |
| Shotgun metagenomics | All the DNA in the sample | Bacteria down to species or strain, plus the genes they carry. More detail, higher price. | Tiny Health (focused on babies, children and pregnancy) |
| Metatranscriptomics | The RNA in the sample | Which microbial genes are active at the time of sampling. | Viome |
ZOE is a different model: the gut test is part of a nutrition program and app, and its offer was reorganized in 2025, so check its current terms and countries before buying. Prices and offers change often for all of these companies: always check the provider's own site.
What the NIST study means for you
NIST sent identical portions of one stool sample to seven consumer services, three kits each. Only 3 genera appeared in every result, three services reported Clostridioides difficile while four did not, and one service called the same sample "healthy" twice and "unhealthy" once. For most common genera, the differences between services were as large as the natural differences between people. The companies were not named, so the study does not say which of the brands above, if any, took part. We explain it in detail in our article on microbiome test reliability.
Practical consequences: compare results only within the same company and method over time, never between two companies; and do not base big diet or supplement decisions on one score.
Five questions to ask before you buy
- Which method? 16S, shotgun or RNA, as above. More detail is not the same as more accuracy.
- Do I get my raw data? A downloadable data file lets you get a second opinion or re-analyze later. Biomesight, for example, says it provides FASTQ and CSV downloads; many services only give a web report. Ask before you pay.
- What counts as "healthy"? There is no agreed definition of a healthy microbiome. Each company uses its own reference population and ranges.
- Are the recommendations tied to products they sell? Supplement or probiotic subscriptions are a common part of the business model.
- Can I retest with the same method? The most useful thing a microbiome test can show is how your own result moves over time, measured the same way.
Do you need a microbiome test at all?
For most people, no. The diet changes with the best evidence work without one: eat many different plants (in the American Gut Project, people who ate more than 30 different plants a week had more diverse gut bacteria than those who ate fewer than 10), plenty of fiber, and some fermented foods if you tolerate them. Persistent digestive symptoms call for a doctor, who can order validated tests when needed.
What you can learn free before spending anything
If you have already taken a DNA test (23andMe, AncestryDNA, MyHeritage, FamilyTreeDNA or Living DNA), your raw file holds three gut-related answers that do not change from one kit to the next: your secretor status (FUT2), whether you keep digesting lactose (lactase), and your gut histamine clearance (DAO). Our free secretor status test reads them in your browser. It will not list your bacteria, but it tells you something stable about the environment they live in. See what your DNA can and cannot say about your microbiome.
Fuel Your DNA does not sell microbiome tests and has no affiliate link with the companies named. Sources: NIST, Communications Biology, February 2026; company websites; McDonald et al., mSystems 2018 (American Gut Project).