The 10-20 Electrode System Explained

Someone Had to Agree on Where to Put the Electrodes

In 1929, a German psychiatrist named Hans Berger published a paper that should have changed the world overnight. He had done something no one had ever done before: he recorded electrical activity from a living human brain. He called the technique "electroencephalography," which is just a very German way of saying "writing down what the electricity in your head is doing."

The technique worked. That was the easy part. The hard part was what came next.

Within a decade, laboratories across Europe and North America were strapping electrodes to people's scalps and recording brainwaves. Researchers were generating mountains of data. There was just one enormous problem: nobody could agree on where the electrodes should go.

One lab in London might place an electrode two inches above the left ear. A lab in Boston might place it three inches above. A lab in Paris might put it somewhere else entirely. They were all recording from "the brain," but they were recording from different parts of the brain, and there was no way to compare their results.

Imagine trying to build a map of a continent if every cartographer used a different coordinate system. That's essentially what EEG research looked like for its first three decades. Thousands of recordings. No common language to talk about them.

This problem needed a solution. And in 1958, a Canadian neuroscientist named Herbert Jasper provided one.

Herbert Jasper and the Problem That Almost Broke EEG

Herbert Jasper was not someone who tolerated ambiguity well. As the head of the electroencephalography laboratory at the Montreal Neurological Institute (working alongside the legendary neurosurgeon Wilder Penfield), Jasper spent his days trying to use EEG to locate seizure foci in epilepsy patients. Precise electrode placement was not an academic concern for him. It was the difference between a successful surgery and a catastrophic one.

Jasper understood something that many of his contemporaries had overlooked: the human skull is not a uniform sphere. Heads come in different sizes and shapes. A position that sits over the motor cortex on one person's head might sit over the parietal association cortex on another's. You cannot just measure a fixed distance in centimeters from some landmark and call it a day, because that fixed distance represents a different proportion of the skull on different people.

So Jasper proposed something elegant. Instead of using absolute measurements, use proportional ones. Measure the total distance between anatomical landmarks on the skull, then divide those distances into percentages. Specifically: 10% and 20%.

The International Federation of Clinical Neurophysiology adopted Jasper's system that same year. They called it the International 10-20 System. And just like that, EEG researchers around the world finally had a common language.

Here's the part that should make you pause: we are still using this system nearly 70 years later. Not because no one has tried to improve it. They have. But because Jasper got the fundamentals so right in 1958 that every subsequent system (the 10-10, the 10-5, the various high-density extensions) is built on top of his original framework rather than replacing it.

The Anatomy of the System: Landmarks, Lines, and Percentages

To understand how the 10-20 system works, you need to know three things about your skull.

The Landmarks

Every measurement in the 10-20 system starts from anatomical landmarks, bony reference points that exist on every human skull regardless of size or shape.

The nasion is the dip at the bridge of your nose, right where the bone of your forehead meets your nasal bones. You can feel it if you run a finger down from your forehead toward your nose. There's a clear indentation. That's the nasion.

The inion is the bump at the back of your skull, roughly where your head curves inward to meet your neck. Run your hand up from your neck along the midline of your skull, and you'll feel a small protuberance. That's the inion.

The preauricular points are the small depressions just in front of each ear, right above the little cartilage flap (called the tragus) that partially covers your ear canal. Press in front of your tragus and open your jaw. You'll feel a small gap where the jawbone moves. That's the preauricular point.

These four landmarks (nasion, inion, left preauricular, right preauricular) define two fundamental lines on your skull that serve as the scaffolding for everything else.

The Measurement Lines

The anteroposterior line runs from the nasion, over the top of your head, to the inion. This line runs front-to-back along the midline of your skull.

The coronal line runs from the left preauricular point, over the top of your head, to the right preauricular point. This line runs left-to-right.

These two lines cross at the very top of your skull. That intersection point is called Cz (the "C" stands for central, and the "z" stands for zero, meaning midline). Cz is the anchor point for the entire system.

The 10% and 20% Intervals

Here's where Jasper's insight comes in. Once you've measured the total distance along the anteroposterior line (nasion to inion), you divide it into segments using percentages:

• 10% from the nasion gives you the Fpz position (frontopolar, midline)
• 20% further gives you Fz (frontal, midline)
• 20% further gives you Cz (central, midline, at the top of the head)
• 20% further gives you Pz (parietal, midline)
• 20% further gives you Oz (occipital, midline)
• The remaining 10% brings you to the inion

The same principle applies along the coronal line and along circumferential lines around the head. The outermost positions always sit at 10% from the landmarks, and the inner positions are spaced at 20% intervals.

This proportional approach is what makes the system universal. A person with a large head has a longer nasion-to-inion distance than someone with a small head. But 20% of a longer distance is a proportionally longer segment, which means the electrode still lands over the same region of cortex. The brain scales with the skull, and the 10-20 system scales with both.

What Do F3, Cz, and O1 Actually Mean?

Every electrode position in the 10-20 system gets a label made up of one or two letters followed by a number (or the letter "z"). Once you understand the naming convention, you can decode any position instantly.

The Letters: Brain Regions

Some extended positions combine letters: CP means centroparietal (between central and parietal), PO means parieto-occipital (between parietal and occipital), FC means frontocentral, and so on.

The Numbers: Hemispheres and Distance from Midline

The numbering system is wonderfully logical once you see it:

Odd numbers (1, 3, 5, 7) are on the left hemisphere. Even numbers (2, 4, 6, 8) are on the right hemisphere. The letter "z" (standing for zero) marks the midline.

Lower numbers are closer to the midline. Higher numbers are farther from it, closer to the temporal regions.

So F3 means: frontal region, left hemisphere, relatively close to the midline. F4 means: frontal region, right hemisphere, same distance from midline as F3. T7 means: temporal region, left hemisphere, far from the midline (literally over the temple). Cz means: central region, dead center on the midline, at the very top of the head.

This encoding system is so compact and expressive that when a neuroscientist says "we saw a spike at P4," every other neuroscientist in the world knows exactly where that electrode was sitting: right parietal region, about 40% of the way from the midline to the right preauricular point. No ambiguity. No room for confusion.

That's the power of a good standard.

The 21 Positions: A Map of the Original System

The original 10-20 system defines 21 electrode positions. Here they are, organized by region:

You might have noticed something: positions like T3, T4, T5, and T6 appear in parentheses next to newer labels. The original 1958 system used the older naming convention. When the American Clinical Neurophysiology Society updated the system in 1991, they renamed these positions to fit a more consistent numbering scheme. You'll encounter both sets of labels in the literature, so it's worth knowing that T7 and T3 refer to the same physical location.

The Neurosity Crown gives you real-time access to your own brainwave data across 8 EEG channels at 256Hz, with on-device processing and open SDKs.

See the Crown

Why Does EEG Standardization Matter So Much?

At this point you might be thinking: okay, it's a coordinate system for the scalp. That's useful, sure, but is it really that important?

Yes. It is arguably one of the most important things that has ever happened to brain science. And here's why.

Reproducibility: The Foundation of Science

Science runs on replication. If a lab in Tokyo discovers that alpha brainwaves increase over the parietal cortex during meditation, a lab in Berlin should be able to place electrodes at the same positions, run the same experiment, and see the same result. Without the 10-20 system, "the same positions" would be meaningless. Every replication attempt would be an apples-to-oranges comparison.

The 10-20 system didn't just standardize electrode placement. It standardized an entire field's ability to have a conversation about what it was finding.

Clinical Diagnosis

Neurologists use EEG to diagnose epilepsy, monitor brain injuries, detect sleep disorders, and assess brain death. In all of these applications, the specific location of abnormal activity matters enormously. An epileptic spike at F7 suggests a different seizure focus (and requires a different surgical approach) than a spike at T8. The 10-20 system ensures that when a neurologist in Mumbai writes "epileptiform discharge at F7-T7" in a clinical report, a neurologist in Chicago knows exactly which brain region is involved.

Brain-Computer Interfaces

And this is where the story connects to the present. Every brain-computer interface, whether it's a 256-channel research system or a consumer headset, uses the 10-20 system to specify where its sensors make contact with your head. This matters because different brain regions produce different signals, and the placement of your electrodes determines what information you can extract.

Want to detect motor imagery (imagining moving your hands) for a BCI? You need electrodes over C3 and C4, the positions overlying the motor cortex that controls the hands. Want to measure visual attention? You need occipital electrodes at O1, Oz, and O2. Want to track frontal alpha asymmetry for emotional regulation research? You need F3 and F4.

The 10-20 system is what makes it possible to design a BCI with confidence that your sensors are actually over the brain regions you care about.

From 21 Electrodes to 256: The Extended Systems

Jasper's original 21-position system was designed for the technology of 1958. As EEG amplifiers got smaller, cheaper, and more capable, researchers wanted more electrodes for finer spatial resolution.

The first major extension, the 10-10 system (sometimes called the 10-20 extended or the SI system), was formalized in 1991. It fills in the gaps between the original 10-20 positions, creating a denser grid with positions at every 10% interval rather than alternating between 10% and 20%. This gives you 81 positions instead of 21.

The naming convention extends naturally. Positions between F3 and C3 become FC3 (frontocentral) and FC1. Positions between C3 and P3 become CP3 (centroparietal) and CP1. The letter-number logic stays the same: combined region letters, odd numbers left, even numbers right, z for midline.

The 10-5 system, published in 2007, pushes even further, defining over 300 positions by filling in 5% intervals. This level of density is used with high-density EEG systems (128 or 256 channels) that researchers use for detailed source localization, essentially trying to reconstruct the three-dimensional electrical activity inside the brain from the two-dimensional map of voltages on the scalp.

Here's the "I had no idea" moment: the spatial resolution of EEG is fundamentally limited not by how many electrodes you use, but by the physics of volume conduction. Electrical signals from the brain spread out as they pass through cerebrospinal fluid, skull, and scalp. By the time they reach the surface, a signal that originated from a focal source in the cortex has been smeared across several centimeters. This means that beyond about 64 to 128 electrodes, adding more channels provides almost no additional spatial information. The signal itself is already blurred.

This is why strategic placement of a smaller number of electrodes at well-chosen positions can be remarkably effective. You don't need to measure everywhere. You need to measure at the right places.

How Consumer Devices Choose Their Subset

Research-grade EEG systems can afford to use 64, 128, or even 256 channels because they're stationary lab equipment connected to high-powered amplifiers. Consumer EEG devices need to be portable, comfortable, and practical. That means fewer channels. And fewer channels means every electrode position counts.

The art of designing a consumer EEG device is choosing which subset of 10-20 positions gives you the most useful information for your intended application.

A device designed purely for meditation feedback might focus on frontal and parietal midline positions (Fz, Pz) to capture the alpha and theta activity associated with meditative states. A device designed for sleep tracking might prioritize frontal positions (where sleep-specific EEG signatures like K-complexes and sleep spindles and K-complexes are most prominent). A device designed for motor imagery BCI would need C3 and C4 at minimum.

But what if you wanted to build a device that could do all of these things? That could detect focus, track meditation, support developer-built applications, and serve as a general-purpose brain-computer interface?

You'd need to cover as many brain regions as possible with as few electrodes as you can get away with. And you'd need to choose positions that capture the widest range of useful signals.

Eight Channels, Four Lobes: How the Neurosity Crown Uses the 10-20 System

The Neurosity Crown places its 8 EEG channels at these positions: CP3, C3, F5, PO3, PO4, F6, C4, and CP4.

If you've been following the naming convention, you can decode these instantly:

Notice the symmetry. Four electrodes on the left hemisphere, four on the right. This bilateral coverage is essential for measuring asymmetry patterns, like the frontal alpha asymmetry that's a key marker for emotional regulation and approach/withdrawal behavior.

Notice the spread. The positions span from frontal (F5/F6) to parieto-occipital (PO3/PO4), covering all four lobes of the cerebral cortex. This means the Crown can pick up signals related to executive function (frontal), motor imagery and somatosensory processing (central), sensory integration (centroparietal), and visual processing (parieto-occipital).

Each of these 8 channels samples at 256Hz, taking 256 measurements per second. That's 2,048 data points per second flowing from your brain to the Crown's onboard N3 processor. The N3 chipset handles signal processing directly on the device, which means your raw brainwave data never needs to leave your head unless you explicitly tell it to.

For developers building on the Neurosity platform, these standardized positions matter because they determine what kinds of applications you can create. The C3/C4 coverage enables motor imagery classification. The frontal positions support attention and focus monitoring. The parieto-occipital positions capture the alpha and beta rhythms that underlie states of calm, visual engagement, and meditative absorption. The Neurosity SDK gives you access to raw EEG data at 256Hz from all 8 channels, plus computed metrics like power spectral density, focus scores, and calm scores.

Because every position maps to the 10-20 system, data from the Crown is directly comparable to data from any other EEG device using the same positions. A machine learning model trained on C3/C4 data from a research-grade system can, in principle, be adapted to work with C3/C4 data from the Crown. The 10-20 system makes that cross-device compatibility possible.

Placing Electrodes in Practice: The Physical Measurement

If you've ever watched someone set up a clinical EEG, it looks almost ritualistic. The technician takes a flexible measuring tape, places one end at the nasion, stretches it over the top of the head to the inion, notes the total distance, then begins marking positions on the scalp with a grease pencil at 10% and 20% intervals.

Here's the step-by-step process for the anteroposterior midline:

1. Measure the total distance from nasion to inion along the midline of the skull. Let's say it's 36 centimeters.
2. Calculate 10% of 36 cm = 3.6 cm. Mark this point above the nasion. This is Fpz.
3. Calculate 20% of 36 cm = 7.2 cm. Mark this distance above Fpz. This is Fz.
4. Mark another 7.2 cm above Fz. This is Cz.
5. Mark another 7.2 cm above Cz. This is Pz.
6. Mark another 7.2 cm above Pz. This is Oz.
7. The remaining 3.6 cm (10%) brings you to the inion. Measurement verified.

The same process repeats along the coronal line (left preauricular to right preauricular, passing through Cz) to establish the lateral positions. Then circumferential measurements at different anterior-posterior levels fill in the rest of the grid.

For consumer devices like the Crown, this manual measurement process isn't necessary. The device is designed so that when you place it on your head in its natural resting position, the electrodes automatically sit at their intended 10-20 positions. The industrial design does the measurement for you. But knowing what those positions mean, and why they're where they are, gives you a much deeper understanding of what your device is actually measuring.

What the 10-20 System Cannot Tell You

The 10-20 system is brilliant, but it has real limitations worth understanding.

It maps the scalp, not the brain. The 10-20 system tells you where an electrode sits on the surface of your head. It doesn't tell you exactly which brain structure is generating the signal that electrode picks up. Due to volume conduction (the way electrical signals spread through biological tissue), an electrode at C3 picks up activity not just from the cortex directly beneath it, but from a broader region. The correspondence between scalp position and underlying brain structure is approximate, not precise.

Brains vary more than skulls. While skull shape is fairly consistent across people (which is why the proportional system works so well), the folding patterns of the cortex underneath are not. The exact location of, say, the hand area of the motor cortex varies by up to a centimeter between individuals. So C3 is over the hand motor cortex in most people, but not all.

It only covers the cortex. EEG, by its nature, primarily detects activity from the cerebral cortex, the wrinkled outer layer of the brain. Deep brain structures like the hippocampus, amygdala, and basal ganglia generate electrical activity too, but their signals are too weak and too far from the scalp to be reliably detected by surface electrodes. The 10-20 system maps the surface because that's what surface electrodes can see.

These limitations are real, but they don't diminish the system's value. For the vast majority of EEG applications, from clinical diagnosis to BCI to neurofeedback, the cortical activity captured at 10-20 positions is rich with useful information. Knowing the limitations just helps you interpret that information more accurately.

A System Built for Humans, Used by Machines

There's something quietly remarkable about the 10-20 system that's worth stepping back to appreciate.

Herbert Jasper designed it in 1958, before transistors were cheap, before digital computers were common in laboratories, before anyone had heard of a brain-computer interface. He designed it to solve a very human problem: getting scientists to agree on a common reference frame so they could share their work.

Nearly seven decades later, that same system is the foundation for technology Jasper could never have imagined. When you put on a Neurosity Crown and a machine learning algorithm classifies your brain state in real time, it's working because each of those 8 sensors sits at a position that Jasper's system defined. When a developer builds an application using the Neurosity SDK that responds to your focus level or detects when you're trying to move an object with your mind, every data point is labeled with a 10-20 position code.

The system has outlasted the vacuum tube era, the transistor era, and the digital revolution. It's now functioning as infrastructure for the brain-computer interface era.

And it works because Jasper understood something fundamental: before you can build anything on top of data, you need to know exactly where that data came from. The 10-20 system is, at its core, a system of trust. When someone says "C3," you know what they mean. When a device says it records from PO3, you know what brain region it's covering. That shared understanding is what makes everything else possible.

Your brain produces electrical signals every waking (and sleeping) moment. The 10-20 system is the Rosetta Stone that lets us read them.