Which Stool Test Is Actually Worth Paying For?

If you have paid several hundred dollars for a stool test and come away more confused than before, you are not alone. One report flags a dozen organisms you have never heard of. Another says everything looks normal while you still feel unwell. Much of that confusion comes from the method behind the test. Stool tests use very different technologies, and each one answers a different question. Once you know which question a test answers, you can tell whether its result is worth acting on.

I first wrote about this in 2019, when I was confused myself by the number of labs and the claims each one made. The technology has moved on since then, a few things I wrote then were wrong, and there is now much better evidence on how reliable these tests are. This is the updated version.

What should a stool test actually tell you?

When you pay for an expensive test, you want two things: a reason for why you feel the way you do, and a clear next step. A practitioner needs a few more:

• Can I interpret it? A list of organisms means little without knowing what they do.
• Does it match the symptoms? A finding that has nothing to do with how the person feels is a distraction.
• Is it reliable? If the same sample were tested again, would it give the same answer?
• Is it actionable? Will the result change what we do?
• Does it show how digestion is working? Inflammation, fat digestion and pancreatic enzyme output often matter as much as which bacteria are present.

No single method ticks every box. Each has strengths and blind spots, and the right test depends on what you need to find out.

The main stool testing methods compared

Culture: growing what is alive

Culture is the oldest method and is still the standard in hospital microbiology. The lab spreads a stool sample on growth plates and waits to see what grows. Because only living organisms grow, a positive culture tells you the organism was alive in your gut, not just leaving traces behind. Once an organism is growing, the lab can also test which agents stop it. This is called susceptibility testing, and it is the main reason some practitioners prefer culture.

Many labs now identify the colonies they grow with MALDI-TOF, a technology that hits a colony with a laser and reads the protein fingerprint it gives off. That fingerprint is matched against a library of known organisms. It was developed in the late 1980s and the first system was cleared for clinical use in the US in 2013. It is fast and accurate for organisms that are in its library, though less so for the many gut bacteria that labs have rarely seen.¹

The weakness is what culture cannot grow. Most bacteria in the large bowel are strict anaerobes, meaning they die when exposed to oxygen. Research labs have shown that a large share of them can be grown with specialised methods,² but a routine stool culture still misses most of the community. Culture also takes days. And when labs test herbs or natural agents against a cultured organism, that test happens in a dish. Whether a result in a dish predicts what happens in your gut has not been shown.

PCR: searching for specific DNA

PCR (polymerase chain reaction) finds a short stretch of DNA that belongs to a specific organism and copies it over and over until there is enough to detect. Real-time PCR (also called quantitative PCR or qPCR) watches the copies build up as it goes, which lets the lab estimate how much of that organism’s DNA was in the sample. Multiplex PCR runs many of these searches at once, which is how one test can check for dozens of organisms.

One correction to what I wrote in 2019. I said PCR was like getting a person’s full name, address and phone number because it reads the whole DNA strand. That was wrong. Targeted PCR does not read an organism’s genome. It looks for one pre-chosen stretch and answers a yes-or-no question: is this particular organism here? If an organism is not on the lab’s list, the test cannot see it. Panels have grown a lot since then, and many now include organisms such as Klebsiella and Proteus that were hard to test this way a few years ago.

PCR is very good at finding known pathogens and parasites quickly. Its limits are worth understanding before you act on a result:

• It counts DNA from dead organisms. DNA lingers after an organism dies, and DNA-based methods cannot reliably tell living cells from dead ones.³ A positive result can reflect something your immune system already dealt with.
• Its “counts” are estimates. Some labs report PCR results in units that look like colony counts. These are calculated from DNA copy numbers. Nothing was grown.
• Resistance genes are not the same as a resistant infection. PCR can detect antibiotic resistance genes in people who are otherwise well, sometimes left over from past antibiotic courses. A resistance gene in a stool sample does not mean the organism carrying it is causing your symptoms.

That last point is the one I worry about most. Acting on DNA remnants can mean going after a problem that is not there.

16S sequencing: the bird’s-eye view

All bacteria carry a gene called 16S rRNA, and part of it varies from one type of bacterium to another. 16S sequencing reads that gene across the whole sample and compares the results to a reference library. The result is a picture of which bacteria make up your gut community and in what proportions.

The catch is detail. It is a bit like knowing there is a Mr Smith somewhere in Australia. There are a great many Mr Smiths, and you still do not know which one, where he lives or what he is doing. Most commercial 16S tests read only a short section of the gene, which usually identifies bacteria to the genus (family name) level rather than the species. Reading the full gene can get to species, but most consumer tests do not do that.⁴

16S results are proportions, not amounts. If one group of bacteria falls, everything else looks higher, even if nothing else changed. The method cannot test which agents would work against an organism, and like PCR, it cannot tell living from dead.

Shotgun metagenomic sequencing: reading everything

This method did not appear in my 2019 article because it was rarely available outside research. Shotgun sequencing breaks up all the DNA in a sample and reads it, rather than targeting one gene. It picks up more of the less common bacteria than 16S does, can identify bacteria to the species level, and shows which genes are present. Those genes hint at what the bacteria are capable of doing.⁵

It is the most detailed picture of the gut community available. It is also the most expensive, it still cannot tell living from dead, and nobody has agreed what a “healthy” profile looks like. More detail does not automatically mean more useful.

Will the lab get the same answer twice?

In 2019 I said the first question to ask any lab is whether it gets the same result if the same sample is tested again. The evidence since then makes this question more important, not less.

A stool is not uniform. Different parts of the same bowel motion hold different bacteria, so testing a scoop from one spot can give a different reading than a scoop from another, unless the lab mixes the whole sample first.⁶ That is one reason two tests taken days apart can disagree.

In 2026, researchers blended one stool sample until it was completely uniform and sent it to seven consumer gut microbiome companies, three kits each. The companies disagreed with each other about which bacteria were present and in what amounts. Some disagreed with themselves: one company rated the same sample healthy twice and unhealthy once. The differences between companies were about as large as the natural differences between two different people.⁷

If you are considering a test, ask the lab what its repeatability is. The technical term is the inter-assay coefficient of variation. A lab that cannot answer that question has not checked.

What the evidence says, and what it leaves out

In 2025, a large international panel of gut microbiome researchers published a consensus statement on microbiome testing. Their conclusion was blunt: the evidence that microbiome profiling helps guide care is scarce, consumer tests are largely unregulated and have no proven clinical value, and most health professionals have not been trained to interpret them.⁸ I agree with much of that, and the reproducibility study above supports it.

Two things are worth adding. First, that statement is about profiling the whole bacterial community. It is not about testing for a specific pathogen or parasite, which is long established. Second, many functional stool panels include digestive and inflammation markers alongside the bacteria, and these are among the best-supported parts of the report. Faecal calprotectin, a marker of bowel inflammation, is recommended in the UK to help tell inflammatory bowel disease apart from irritable bowel syndrome.⁹ Pancreatic elastase is widely used to check whether the pancreas is producing enough digestive enzymes. These markers are often more actionable than the organism list.

What the mainstream view tends to underweight is the person with ongoing symptoms whose standard pathology stool test comes back clear. Those routine tests look for a narrow set of pathogens. A clear result rules those out. It does not mean the gut is working well.

So the weight of evidence sits here: testing for known pathogens and parasites, and the digestive and inflammation markers, stand on solid ground. A ranking of “good” and “bad” bacteria, or a score for gut health, does not yet.

Read the person, not the report

This is the part I most want practitioners and clients to hear. No test is 100% accurate. A result is a clue, and it only means something when it fits the person’s symptoms, history and other results. The most common mistake I see is people acting on the numbers on the page rather than on how the person actually feels. A finding that does not match the symptoms is rarely the answer, however alarming it looks.

The goal is never to change numbers on a piece of paper. It is to help someone feel better.

How to choose a stool test

• Looking for an infection or parasite: a targeted PCR panel or culture, ideally ordered with your doctor if you have red-flag symptoms such as blood in the stool, weight loss or fever.
• Wanting to know what would work against a specific organism: culture with susceptibility testing, keeping in mind that herbal susceptibility is a dish result.
• Wanting an overview of the bacterial community: shotgun sequencing gives more detail than 16S, but neither should drive decisions on its own.
• Wanting to know how digestion is working: a panel that includes calprotectin, pancreatic elastase and fat markers.
• Before you pay: ask about repeatability, whether results distinguish living from dead organisms, and how amounts are measured. Be cautious of a report that comes bundled with recommendations for products sold by the same company.

Frequently asked questions

Is culture or PCR better for stool testing?
Neither is better across the board. Culture shows what is alive and can test which agents work against it, but misses most gut bacteria. PCR is faster and more sensitive for known pathogens and parasites, but only finds what it targets and counts DNA from dead organisms. Many practitioners use both for different questions.

Why did my stool test show antibiotic resistance genes when I feel fine?
PCR is sensitive enough to detect resistance genes left in the gut from past antibiotic courses or carried by harmless bacteria. Finding the gene does not mean you have a resistant infection. It only matters if it fits your symptoms and other findings.

Why did my stool test come back normal when I still have symptoms?
Standard pathology stool tests check for a narrow list of pathogens, so a clear result rules those out and nothing more. Symptoms can come from digestion, inflammation, the small intestine or outside the gut entirely. A normal result is still useful information.

Can a stool test diagnose SIBO?
No. A stool sample reflects the large bowel, while SIBO (small intestinal bacterial overgrowth) happens in the small intestine. Breath testing is the usual way to investigate it.

Are at-home gut microbiome tests reliable?
Not reliably enough to base decisions on yet. When researchers sent the same uniform sample to seven consumer companies, results differed between companies and sometimes between kits from the same company. Testing for specific pathogens through a lab is a different matter and is well established.

Should I repeat a stool test to see if things have changed?
Only if the result will change what you do next. Stool results vary from sample to sample and between labs, so if you do retest, use the same lab and method, otherwise you may be comparing the methods rather than your gut.