The endoscopy test for small intestinal bacterial overgrowth

In short: the most direct way to diagnose SIBO is to take a sample of fluid from the small intestine during an upper endoscopy and count the bacteria in a laboratory. More than 1,000 colony-forming unitsColony-forming units (CFU) A count of live microbes in a sample. SIBO is defined as more than 1,000 CFU per milliliter of small-intestinal fluid (older definitions used 100,000). per milliliter (CFU/mL) is the current threshold for SIBO¹. The test is invasive, expensive and can give false results: the sample can be contaminated on the way down, and a single sample can miss overgrowth further along the gut². For most people, a breath test is used instead.

Small-intestinal fluid culture is often called the reference or “gold standard” test for SIBO. This guide explains how it is done, how results are interpreted, what newer sequencing methods add, and why it is not used more often. For an overview of all tests, see our diagnosis guide.

How the test is done

The sample is taken during an upper endoscopy (gastroscopy), in which a flexible camera is passed through the mouth and stomach into the first part of the small intestine (the duodenum), usually under sedation. A 2025 position paper describes a typical technique²:

• The endoscope is passed into the small intestine with minimal air.
• The endoscopist switches to sterile gloves, and a thin sterile catheter is passed through the endoscope.
• Over 3 to 5 minutes, 2 to 5 mL of small-intestinal fluid is gently drawn into a syringe and sent for culture.

Biopsies of the intestinal lining may be taken at the same time, for example to check for celiac disease. They do not diagnose SIBO: in a 2010 study, more than half of the biopsies from people with SIBO looked normal⁹.

Reading the result

The laboratory grows the bacteria on culture plates and counts the colonies.

• Threshold. The traditional threshold for jejunal fluid was 100,000 CFU/mL or more. A threshold of 1,000 CFU/mL or more has since been adopted for duodenal fluid², and the 2020 US guideline uses more than 1,000 CFU/mL¹.
• Why 1,000. In a 2024 study of 483 people undergoing endoscopy, those with 1,000 CFU/mL or more had lower microbial diversity, more Escherichia and Klebsiella, and more severe abdominal pain, diarrhea and bloating. A few strains of E. coli and two Klebsiella species made up about 40% of all duodenal bacteria in these people⁵. A 2020 study of 140 people found the same pattern of increased Proteobacteria, the group that includes E. coli and Klebsiella, in SIBO⁴.

What sequencing adds

Culture only detects bacteria that grow in the laboratory, and only about 40% of gut bacteria can be identified with conventional culture². Research groups now also analyze the DNA in small-intestinal fluid. A 2019 study found that treating samples to break down mucus before analysis recovered almost three times more anaerobic bacteria⁶. These sequencing methods are used in research and are not yet part of routine diagnosis.

Limitations

The test has several drawbacks²:

• It is invasive and costly, requires sedation and is not available everywhere.
• Contamination. Bacteria from the mouth and stomach can enter the sample. In one study, a single-channel catheter gave a contamination rate of 19.6%.
• Sampling. Bacteria are not spread evenly through the small intestine, so a sample from the upper part can miss overgrowth further down.
• Culture limits. Not all hydrogen-producing bacteria can be cultured, and many laboratories cannot culture methane-producing archaeaArchaea Single-celled microorganisms that are distinct from bacteria. Methane-producing archaea (methanogens) are responsible for intestinal methanogen overgrowth., so the test cannot diagnose methane overgrowth.

A US expert review noted that a major barrier to defining SIBO is how little is known about the normal small-intestinal microbiome³, and the European breath test guideline recommends breath testing “until a true gold standard is established”⁷.

Endoscopy vs. breath testing

Compared with fluid culture, the glucose breath test detects about 54% of cases with 83% specificity, and the lactuloseLactulose A synthetic sugar the body cannot digest or absorb. As a breath-test substrate it travels the whole small intestine, which also makes false positives more likely. test about 42% with 71% specificity⁸. In the 2020 culture study, only 4 of 7 people with SIBO on culture had a positive lactulose breath test at 90 minutes, compared with 2 of 13 without SIBO⁴.

When is endoscopy used for SIBO?

In practice, fluid culture tends to be used where it changes management: when breath test results are unclear, when someone is already having an endoscopy for another reason, or in complex cases such as previous gut surgery. Our diagnosis guide explains how to choose between tests.

Common questions

Is endoscopy better than a breath test for SIBO?

It is more direct, but it has its own false positives and negatives and cannot detect methane overgrowth². Most people are diagnosed with a breath test⁷.

Can a normal endoscopy rule out SIBO?

Only if fluid was taken and cultured. A visual exam or biopsy alone cannot diagnose or rule out SIBO, because biopsies often look normal in SIBO⁹.