Eye tracking is a technology that measures where your eyes look, how long they linger, and how they move between points of interest. It works by detecting the position of your pupils and the reflection of light off your eyes, then translating that data into precise coordinates on a screen, object, or environment. What started as a niche research tool has spread into medicine, automotive safety, accessibility, virtual reality, and everyday product design.
How Eye Tracking Works
Most modern eye trackers use infrared light. A small emitter shines near-infrared beams toward your eyes, creating a reflection pattern on the cornea. A camera captures both that reflection and the position of your pupil, and software calculates exactly where your gaze lands. This approach works because the relationship between the corneal reflection and the pupil center shifts predictably as your eyes move.
Before you use an eye tracker, you typically go through a quick calibration step where you follow a dot around the screen. This teaches the system how your specific eye geometry maps to screen coordinates. Once calibrated, the tracker records your gaze continuously, often hundreds of times per second.
Key Measurements Eye Trackers Capture
Eye trackers don’t just record “where you looked.” They break your gaze behavior into distinct components that reveal how you process information. The three most important are fixations, saccades, and dwell time.
Fixations are the moments when your eyes pause on something. Despite what it feels like, your eyes don’t sweep smoothly across a page or screen. They jump from point to point, and each pause is a fixation, typically lasting a few hundred milliseconds. Fixations indicate what caught your attention and what your brain is actively processing.
Saccades are the rapid jumps between fixations. They happen so fast you’re not consciously aware of them. The pattern of saccades reveals your scanning strategy: whether you read every line of text, skimmed headings, or jumped straight to an image.
Dwell time measures the total duration your gaze stays in a defined area. Research has shown that dwell time reliably predicts decisions. When people choose between two options, they spend more total time looking at the one they eventually pick, even before they’re consciously aware of their preference.
Hardware: Research Grade vs. Consumer
Eye trackers range from laboratory instruments costing thousands of dollars to software that runs on a standard laptop webcam. The differences come down to speed, accuracy, and what kinds of eye movements they can detect.
Research-grade stationary trackers sample at 300 to 1,000 Hz, meaning they capture your eye position 300 to 1,000 times per second. A high-end system like the EyeLink 1000 records at 1,000 Hz and achieves gaze accuracy around 0.4 to 0.9 degrees of visual angle. For context, one degree of visual angle is roughly the width of your thumb held at arm’s length. These systems can detect the fastest eye movements, including tiny involuntary microsaccades.
Mobile eye trackers, the kind built into lightweight glasses, typically run at 60 to 120 Hz. They sacrifice some speed for the ability to track gaze in real-world environments: a grocery store, a car, a factory floor. At 60 Hz, detecting individual saccades becomes unreliable, but fixation patterns and overall dwell time remain accurate.
Webcam-based eye tracking has improved dramatically. Recent systems achieve accuracy of about 1.4 degrees, compared to roughly 0.9 degrees for a dedicated infrared tracker. That gap of about half a degree matters less than you might think for many applications. The bigger limitation is speed: webcams capture at around 30 Hz, which makes them unsuitable for studying rapid eye movements but perfectly adequate for tracking which part of a webpage someone looks at or how they navigate a product interface.
Eye Tracking in UX and Design
User experience researchers were among the earliest commercial adopters of eye tracking. The core question is simple: when someone visits a website, opens an app, or looks at a product package, where does their gaze actually go?
Two visualizations dominate this field. Heatmaps overlay color gradients on an image of the interface, with warm colors (red, orange) showing where people looked most and cool colors showing areas that were ignored. They’re immediately intuitive and often reveal that elements designers assumed were prominent are being overlooked entirely. Gaze plots trace the sequential path of fixations, showing the order in which someone noticed things and how their eyes moved through a layout. Together, these tools turn subjective opinions about design into measurable behavior.
Medical and Diagnostic Uses
Eye movements are controlled by some of the same brain circuits involved in attention, decision-making, and motor planning. That makes abnormal eye movement patterns a potential window into neurological conditions.
In pediatric concussion assessment, eye tracking has shown promise as a fast, noninvasive screening tool. Children with concussions often develop convergence insufficiency, where the eyes struggle to focus together on a near object, and accommodation dysfunction, where focusing at different distances becomes impaired. Traditional concussion evaluation relies heavily on self-reported symptoms, which is especially unreliable in children. Eye tracking can objectively detect these gaze abnormalities, offering clinicians measurable data rather than relying on a child’s ability to describe how they feel.
Researchers have also explored eye tracking in autism spectrum disorder screening, since differences in social gaze, particularly reduced attention to faces and eyes, appear early in development. While no eye tracking test has replaced standard diagnostic evaluation, the ability to quantify gaze patterns in preverbal infants opens possibilities that traditional behavioral observation cannot.
Accessibility and Communication
For people who lose the ability to speak or move, eye tracking can become their primary connection to the world. This is especially critical in amyotrophic lateral sclerosis (ALS), a progressive disease that gradually paralyzes voluntary muscles while leaving cognition intact. Many people with ALS eventually cannot speak, type, or gesture.
Eye-gaze communication systems use cameras to track pupil movement along horizontal and vertical axes, letting someone select letters, words, or icons on a screen just by looking at them. In research on digital communication tools for ALS, eye-gaze interaction was the most common approach, used in over a third of studies. These systems rely on infrared light to create reference points on the pupil and iris, making detection reliable even in varying lighting conditions. The result is that someone who cannot move their limbs can write messages, control a computer, browse the internet, and maintain social connections.
Automotive Safety and Driver Monitoring
Drowsy driving contributes to a significant share of serious crashes. Studies using eye tracking have found that drowsiness, measured by the percentage of time a driver’s eyelids cover more than 80% of the eye area over a one-minute window (a metric called PERCLOS), was present in roughly 10.6 to 10.8% of crashes causing property damage or injury.
Modern driver monitoring systems use cameras pointed at the driver’s face to continuously measure eye closure, blink rate, and gaze direction. Blink rate is simply the number of blinks per minute, which increases with fatigue. The eye aspect ratio, a measure of how open or closed the eye appears, provides a frame-by-frame signal that algorithms use to detect both slow, drowsy blinks and full microsleeps. Some systems also track head pose and mouth movements to detect yawning. When these signals cross a threshold, the vehicle can issue an alert or, in advanced systems, begin automated safety interventions.
Eye Tracking in Virtual Reality
VR headsets face a fundamental performance challenge: rendering high-resolution graphics across your entire field of view is extremely demanding on the processor. Eye tracking solves this through a technique called foveated rendering. Since human vision is only sharp in a small central area (the fovea) and blurry in the periphery, the headset tracks where you’re looking and renders full detail only in that spot. Everything outside your direct gaze is rendered at lower resolution.
Meta’s implementation of this in Quest headsets reduces GPU workload substantially without a perceptible drop in visual quality. Compared to fixed foveated rendering, which always reduces detail in the same peripheral zones regardless of where you look, eye-tracked foveated rendering adapts in real time to your gaze. The practical result is smoother performance, longer battery life in standalone headsets, and the ability to push higher graphical fidelity where it matters most.