Everywhere we go, we leave little traces in the form of the bacteria we carry and the skin and hair we shed. In some cases, it’s even possible to follow these bacterial breadcrumbs back to the person who left them. But that’s not the only way we leave our microbial mark. We also emit what is called a “microbial cloud,” and researchers have found that it is also unique enough to be able to identify individual people.

“We expected that we would be able to detect the human microbiome in the air around a person,” explained James Meadow, who led the study published in PeerJ, “but we were surprised to find that we could identify most of the occupants just by sampling their microbial cloud.” They found that participants of the experiment could be identified within four hours by analyzing the particulates suspended in the air, and looking at the unique combination of bacteria present.

We emit this microbial cloud from our breath, but it’s also made up of the bacteria found on our clothes, skin and hair. Previous studies have already found that the microbes present in house dust can give all sorts of information about those who live there, while others have found that our microbiome can also act like a “fingerprint” in identification. It’s long been known that we must breathe out plenty of bacteria, but no study has been able to show that the cloud we emit is detectable, or whether such clouds could be sufficiently different from person to person to allow identification.

The researchers from the University of Oregon got 11 participants to sit in a sanitized experimental chamber. This filtered the air going in and out so that all bacteria and particulates emitted by the subjects could be trapped. They also placed various surfaces and petri dishes around the participants to catch settled particles. The researchers then took samples and analyzed the results, sequencing the DNA of the microbes present.

From a four hour period, they reported over 14 million sequences representing thousands of different types of bacteria emitted from the participants. Further analysis revealed that “samples from each individual were statistically distinct and identifiable to that occupant.” For example, one subject had significant levels of the bacteria Dolosigranulum pigrum present in their cloud, whereas this was absent from others, while another’s cloud was dominated by Streptococcus.

While it might be unsurprising that people leave this microbial signature wherever they go, the researchers were not expecting such a personalized and unique airborne emission. They hope that this could be used to model the spread of infectious diseases in a built-up environment, and possibly in the field of forensics.