The ethmoid bone is a small, unpaired bone nestled deep in the center of the skull between the eyes, and despite its modest size it plays an outsized role in breathing, smelling, and protecting the brain. It forms much of the upper nasal cavity, contributes to the nasal septum, and makes up the thin inner wall of each eye socket. Most people never think about it until something goes wrong with their sinuses or until a surgeon needs to navigate through it, but the ethmoid is one of the most architecturally complex bones in the human body.
Where It Sits and What It Looks Like
If you could peer straight down through the top of the skull, the ethmoid would be a roughly rectangular block lodged between the two eye sockets, just behind the bridge of the nose. It connects to more neighboring bones than almost any other cranial bone: the frontal bone above, the sphenoid behind, the maxillae and palatine bones below, the lacrimal bones on each side, the nasal bones in front, and the vomer below its midline plate. That web of connections is one reason it can be so clinically consequential when fractured or eroded by disease.
The bone has four main parts worth knowing. At the top is a horizontal shelf called the cribriform plate, perforated with tiny holes like a colander. Running down the midline is the perpendicular plate, which forms the upper portion of the nasal septum. Hanging on each side are two masses of paper-thin, scroll-shaped projections called the superior and middle nasal conchae (also known as turbinates). And filling the space within those lateral masses is a honeycomb of air-filled chambers: the ethmoid sinuses, sometimes called ethmoid air cells.
The Gateway for Smell
The cribriform plate is the structure that links your nose to your brain. Dozens of tiny nerve bundles, collectively forming the olfactory nerve, thread upward through its perforations from the lining of the nasal cavity to the olfactory bulb sitting just above. This is the only place in the body where nerve fibers pass directly from an open cavity through bone into the brain, which makes the cribriform plate both a marvel and a vulnerability.
Recent imaging work has started to reveal the fine architecture of these nerve pathways in living humans. MRI tractography has shown that nerve fibers originating from the nasal septum side project to the inner half of the olfactory bulb, while fibers from the turbinate side project to the outer half, preserving an anterior-to-posterior spatial map along the way. In other words, the brain keeps a rough geographic record of where in the nose an odor molecule landed, which may help explain how you can sometimes sense the direction a smell is coming from.
The Ethmoid Sinuses and Their Surprising Variability
The ethmoid sinuses are not a single hollow chamber like the frontal or sphenoid sinus. They are a cluster of anywhere from three to eighteen small air cells packed into the lateral masses, divided into anterior, middle, and posterior groups. Their mature volume averages around 5,000 cubic millimeters per side, which is modest compared to the maxillary sinuses but still enough internal surface area to cause real trouble when infected or inflamed.
What makes these sinuses particularly interesting to surgeons is how wildly their anatomy varies from person to person. A CT-based study in a Southern Chinese population found that virtually every patient (over 99%) had at least one named anatomical variant of the ethmoid sinuses. The most common were agger nasi cells, present in about 96% of patients, followed by Onodi cells in roughly 60% and Haller cells in about 29%.
The prevalence of specific variants shifts across populations and studies. Work using high-resolution CT in an Indian cohort found Haller cells and Onodi cells each in about 30% of patients, while a separate CT study found agger nasi cells to be the most frequent variant overall, with Onodi cells among the least common. The discrepancies partly reflect genuine population differences and partly reflect how aggressively radiologists look for the variants. What holds constant across all studies is that some form of ethmoid variation is the rule, not the exception.
These variants matter practically because each one changes the surgical landscape. An Onodi cell is an ethmoid air cell that has pneumatized (expanded with air) posteriorly to wrap around the optic nerve. A Haller cell sits beneath the orbit floor, narrowing the drainage pathway of the maxillary sinus. A surgeon who does not identify these variants on a preoperative CT scan risks damaging the optic nerve, the orbit, or the skull base.
The Paper-Thin Wall Between Nose and Eye
The outer wall of each ethmoid labyrinth is a wafer of bone called the lamina papyracea, literally “paper-like plate.” It forms most of the medial wall of the eye socket, and in some people it is so thin it is essentially translucent. This matters because infection or inflammation in the ethmoid sinuses can spread directly into the orbit, causing swelling, vision changes, or in severe cases an orbital abscess.
Even without disease, the lamina papyracea is not always intact. A CT study of 1,000 patients found that about 2% had a spontaneous dehiscence, meaning a gap in the bone through which orbital tissue had herniated into the sinus cavity. In most of those cases the herniated tissue was orbital fat, but in a small number the medial rectus muscle (the muscle that moves the eye inward) had pushed through. During endoscopic sinus surgery, mistaking a bulge of orbital fat for a polyp and cutting into it can cause bleeding, double vision, or worse. Preoperative imaging specifically to check for lamina papyracea gaps has become a standard precaution.
Blood Supply and Nosebleeds
The ethmoid bone’s blood supply comes from two anterior and posterior ethmoidal arteries, which are branches of the ophthalmic artery (itself a branch of the internal carotid). These arteries enter through small canals in the upper orbit, cross the ethmoid roof, and then descend into the nasal cavity, where they supply much of the upper nasal lining.
The anterior ethmoidal artery is particularly relevant to nosebleeds. It feeds into Kiesselbach’s plexus, an anastomotic tangle of small vessels on the front of the nasal septum where most common nosebleeds originate. Cadaver dissections have confirmed that the posterior septal artery, the greater palatine artery, the superior labial artery, and the anterior ethmoidal artery all converge at this plexus. If you have ever pinched the soft part of your nose to stop a nosebleed, you were compressing this network. For posterior nosebleeds that do not stop with pressure, the ethmoidal arteries sometimes need to be clipped surgically, a procedure that requires navigating the ethmoid roof from the orbit side.
Why Sinus Surgeons Worry About the Ethmoid Roof
Endoscopic sinus surgery often involves opening and clearing the ethmoid air cells to improve drainage. The ceiling of the ethmoid labyrinth is the fovea ethmoidalis, which forms part of the anterior cranial base. On either side of the crista galli (the midline crest of the ethmoid), the lateral lamella of the cribriform plate drops down to meet the fovea ethmoidalis, creating the olfactory fossa where the olfactory bulb sits.
In the 1950s, a classification system was developed to describe how deep this olfactory fossa is. A shallow fossa (Keros type I) means the lateral lamella is short and relatively sturdy. A deep fossa (Keros type III) means the lateral lamella is long, thin, and dangerously easy to breach. A surgeon working in a deep olfactory fossa is operating millimeters from the brain, with the thinnest bone in the skull as the only barrier. Punching through it can cause a cerebrospinal fluid leak, introduce bacteria leading to meningitis, or damage blood vessels inside the skull.
Adding to the difficulty, the depth of the olfactory fossa frequently differs between the left and right sides in the same patient. Studies have found meaningful side-to-side asymmetry to be common, which means a surgeon cannot assume that what is true on one side holds on the other. This is why preoperative CT imaging with careful attention to the Keros classification on each side has become essential before any endoscopic procedure near the ethmoid roof.
Cerebrospinal Fluid Leaks and the Cribriform Plate
Because the cribriform plate is so thin and perforated, it is one of the most common sites for cerebrospinal fluid (CSF) rhinorrhea, a condition where the clear fluid surrounding the brain leaks through the skull base and drips out of the nose. This can happen after head trauma, after surgery, or occasionally spontaneously in people with elevated intracranial pressure.
CSF rhinorrhea is easy to mistake for a runny nose or allergies, and it sometimes goes undiagnosed for months. The danger is that the opening in the skull base creates a direct pathway for bacteria from the nasal cavity to reach the brain. A case report described a nine-year-old girl who was admitted twice for meningitis before anyone identified that she had a traumatic CSF leak through the cribriform plate from an earlier injury. Once the leak was repaired surgically through an intranasal approach, the recurrent infections stopped. The case illustrates a broader point: unexplained recurrent meningitis, especially in someone with a history of facial or head trauma, should prompt investigation for a cribriform plate defect.
How the Ethmoid Grows in Children
The ethmoid sinuses are among the first paranasal sinuses to develop, beginning to pneumatize (fill with air) even before birth. At birth, the ethmoid is still largely cartilaginous, and it ossifies (hardens into bone) gradually over childhood. The septoethmoidal junction, where the nasal septum cartilage meets the ethmoid, develops growth-plate-like structures that mineralize at rates comparable to the growth plates in long bones, contributing to facial growth through much of childhood and adolescence.
Three-dimensional imaging studies tracking sinus volume in children have found that the ethmoid sinuses grow rapidly between ages two and eight, then undergo a second growth spurt between roughly thirteen and sixteen, reaching mature volume around age sixteen. That makes the ethmoid sinuses one of the earlier paranasal sinuses to finish growing; by comparison, the sphenoid and frontal sinuses continue expanding until around age twenty. This growth timeline matters for pediatric ENT specialists because operating on ethmoid sinuses in a young child means working in a smaller space that is still actively changing shape.
The Ethmoid as a Crumple Zone
One of the more surprising roles of the ethmoid bone is protecting the brain during facial impacts. Finite element modeling of head trauma has shown that when the face takes a blow, the nasal cartilages absorb initial energy through deformation and fracture, and then the vomer and ethmoid divert stress into the air-filled ethmoid and sphenoid sinuses. These sinuses function like a car’s crumple zone: their thin walls collapse and absorb energy that would otherwise transmit straight to the brain. The same study concluded that, in its most natural manner, the face protects the brain through this sequential energy-dissipation pathway.
This crumple-zone effect also explains why midface fractures can be dramatic in appearance yet cause relatively little brain injury. The ethmoid and surrounding sinuses sacrifice themselves, so to speak, absorbing force that would otherwise reach the anterior cranial fossa. Of course, this protection has limits. A high-energy impact like a car crash can overwhelm the system, fracturing through the ethmoid roof and into the brain. But for moderate impacts, the architecture of the ethmoid works surprisingly well as a built-in shock absorber.
The Ethmoid in Other Animals
The ethmoid bone is not unique to humans. In most mammals it is far more elaborate than ours, packed with additional scroll-shaped bones called ethmoturbinals that dramatically increase the surface area inside the nasal cavity. Dogs, for instance, have a labyrinth of ethmoturbinals that dwarfs the human version, which is part of why their sense of smell is so much more acute.
Research on felids (cats and their relatives) has revealed that the ethmoturbinals serve double duty. In short-snouted species like domestic cats and cheetahs, the front-most ethmoturbinals appear to have been co-opted for respiratory function, warming and humidifying incoming air, rather than serving the olfactory role they play in longer-snouted relatives. The shift illustrates a tension between two competing needs: conditioning the air you breathe and sampling it for chemical information.
In primates, the first ethmoturbinal has been studied as a marker of evolutionary tradeoffs between smell and body-temperature regulation. Nocturnal primates that rely heavily on scent tend to have proportionally larger first ethmoturbinals than their diurnal (daytime-active) relatives, suggesting the bone’s internal architecture reflects dual adaptations for olfaction and for conserving heat and moisture during breathing. Humans, with our reduced reliance on smell and our relatively flat faces, have some of the most simplified ethmoid architecture among primates, a structural simplification that parallels the shrinking of our olfactory repertoire over evolutionary time.
Common Misconceptions About the Ethmoid
A persistent idea in anatomy courses is that the ethmoid is a purely structural bone, interesting mainly as a landmark. In reality, it plays active roles in airflow dynamics, immune defense (the sinus lining produces mucus and hosts immune cells), and even sensory mapping, as the organized projection of olfactory nerve fibers through the cribriform plate shows. Treating it as inert scaffolding undersells what it does.
Another misconception is that sinus anatomy is mostly standard from person to person, with “variants” being unusual exceptions. The CT data tell the opposite story: having some kind of ethmoid sinus variant is nearly universal, and the specific configuration of air cells is practically a fingerprint. Two people’s ethmoid labyrinths can look remarkably different on a scan, even when both are perfectly healthy. Surgeons who operate in this area regularly say that no two ethmoid dissections are exactly alike, and the imaging literature backs them up.
Finally, the idea that the cribriform plate is well protected from injury deserves pushback. Its position at the base of the anterior cranial fossa, above a cavity that is open to the outside world, makes it one of the most exposed parts of the skull base. A hard blow to the nose or forehead, a fall, or even aggressive nasal instrumentation can fracture it. The consequences of that fracture, from loss of smell to meningitis, are disproportionate to what seems like a minor injury, which is one reason emergency physicians are trained to check for clear nasal drainage after any significant facial trauma.