A Marcus Gunn pupil is a pupil that dilates instead of constricting when a light is swung to it from the other eye, signaling that the affected eye’s optic nerve or retina is transmitting light signals to the brain less effectively than its partner. The formal clinical name is a relative afferent pupillary defect, or RAPD. It is not a disease in itself but one of the most reliable bedside signs that something is wrong along the visual pathway between the eye and the brain, and it can be detected with nothing more than a penlight and a dark room.
What Happens During the Swinging Flashlight Test
The test that reveals a Marcus Gunn pupil is remarkably simple. You sit in a dimly lit room while a clinician swings a bright light back and forth between your eyes, spending two to three seconds on each one. In a healthy pair of eyes, both pupils constrict equally every time the light lands on either side, because the signal from each eye is balanced. When one eye has a weaker afferent signal, the brain perceives the light shone into that eye as dimmer than the light shone into the healthy eye. As the beam swings to the affected side, both pupils paradoxically dilate, as though the light had been taken away rather than simply moved. That dilation is the Marcus Gunn pupil.
The key point is that this test compares one eye against the other. A person with equally damaged optic nerves in both eyes may show no RAPD at all, because neither side is weaker relative to the other. Likewise, someone with perfectly healthy eyes obviously shows no defect. The sign only appears when there is an asymmetry between the two sides of the visual input pathway.
Why the Pupil Responds This Way
Light entering the eye travels along the optic nerve to the brainstem, where it triggers a reflex loop that tells both pupils to constrict. Because the reflex is wired to both sides, shining a light in one eye makes both pupils shrink equally. When the optic nerve on one side is damaged, it sends a weaker “there’s light here” message to the brainstem. The reflex arc still works, but the drive is reduced. So when the flashlight lands on the impaired eye, the brain registers less light overall, and both pupils open up a bit.
Research using brain imaging in patients with visual pathway damage has confirmed that the pupillary light reflex pathway runs close to the optic radiation as it begins in the temporal white matter. Damage to that region is commonly associated with a RAPD, even in patients whose primary problem is further back in the brain rather than in the optic nerve itself.1Neurology. The pupillary light reflex pathway: cytoarchitectonic probabilistic maps in hemianopic patients
The Most Common Causes
A Marcus Gunn pupil shows up whenever there is a significant difference in how well each eye relays visual information to the brain. Optic nerve problems are the leading cause, but the list is broader than many people assume.
- Optic neuritis: Inflammation of the optic nerve, often linked to multiple sclerosis, is one of the classic triggers. Research tracking recovery in patients with demyelinating optic neuritis has shown that pupil responses and visual performance both improve over time, but the RAPD can persist even after vision feels subjectively normal.2PubMed. Recovery of vision and pupil responses in optic neuritis and multiple sclerosis
- Optic nerve compression: Tumors, swelling, or bony fractures around the orbit can press on the optic nerve and produce a RAPD. In traumatic optic neuropathy, the severity of the RAPD correlates strongly with eventual visual outcome.3PubMed Central. Predictive value of visual evoked potentials, relative afferent pupillary defect, and orbital fractures in patients with traumatic optic neuropathy
- Retinal vascular occlusion: A blockage in the central retinal vein can damage the retina enough to weaken its signal. The size of the RAPD is a practical way to tell the two types of retinal vein occlusion apart: in patients with the milder, nonischemic form, the defect is typically small (nearly all under 0.9 log units), while in the more dangerous ischemic form, it is substantially larger (nearly all 1.2 log units or more).4PubMed. Relative afferent pupillary defect in central retinal vein occlusion
- Glaucoma: When glaucoma affects one eye more severely than the other, a RAPD can develop on the worse side. Automated pupillometers can pick this up with high accuracy in asymmetric glaucoma cases.5PubMed Central. Quantification of RAPD by an automated pupillometer in asymmetric glaucoma and its correlation with manual pupillary assessment
- Retinal detachment and extensive retinal disease: Any condition that knocks out a large area of retina in one eye can reduce the total light signal enough to produce a relative defect.
Because optic nerve problems account for the vast majority of clinically significant RAPDs, finding one during a routine exam often triggers imaging of the brain and orbits to rule out compressive or inflammatory causes.
Measuring How Severe It Is
Spotting a Marcus Gunn pupil is binary: either the affected pupil dilates when the light swings to it, or it doesn’t. But clinicians also want to know how big the defect is, because that number influences diagnosis and helps track whether a condition is getting better or worse.
The standard quantitative method involves neutral density filters. A filter of known darkness is placed in front of the healthy eye to weaken the light reaching it, effectively handicapping the stronger side. Filters of increasing density are added until the two pupils respond equally, meaning the defect has been neutralized. The filter strength at that balancing point, measured in log units, is the magnitude of the RAPD.6Survey of Ophthalmology. How to measure the relative afferent pupillary defect Common clinical grades line up well with these measurements: a mild grade corresponds to about 0.6 log units or less, a moderate grade to about 0.6 to 0.9, and more severe grades to 1.2 and above.7PubMed. Correlation between clinical grading and quantification by neutral density filter of relative afferent pupillary defect (RAPD)
The same neutral-density-filter principle is now being adapted for newer technologies. A virtual-reality headset system, for example, can repeat the swinging flashlight test with digitally controlled light levels while tracking the pupil with high-speed cameras, removing some of the subjectivity of the bedside exam.8Translational Vision Science & Technology. Detection of Relative Afferent Pupillary Defects Using Eye Tracking and a VR Headset
The Cataract Twist
One of the more counterintuitive findings in pupil testing is that a dense cataract in one eye can create a RAPD in the other, apparently healthy eye. This seems backwards at first: the cataract blocks light, so you would expect the defect on the cataract side. What actually happens is that the cataract scatters and diffuses the light entering that eye, which paradoxically makes the light stimulus more effective at driving the pupil reflex. The healthy eye, receiving a sharper but less broadly scattered beam, ends up looking weaker by comparison.
This effect has been documented in patients with mature and nuclear cataracts, where the contralateral eye showed a measurable RAPD averaging about 0.44 log units. After the cataract was surgically removed, the defect vanished.9PubMed. A unilateral cataract produces a relative afferent pupillary defect in the contralateral eye Pupillography in a separate case confirmed the same pattern quantitatively: a mature cataract in one eye, a RAPD in the other, and complete resolution after cataract surgery.10Canadian Journal of Ophthalmology. Pupillography of relative afferent pupillary defect contralateral to monocular mature cataracta
The practical takeaway matters: if you find a RAPD in an eye that also has a dense cataract on the opposite side, it does not necessarily mean the eye with the defect has a damaged visual pathway. The cataract itself may be the culprit. But the reverse is also true. If a patient with a known optic nerve problem develops a cataract in the same eye, the cataract can mask or even reverse a preexisting RAPD, making the nerve damage look better than it really is. Clinicians have to think carefully about what the cataract is doing to the light stimulus before drawing conclusions.
Amblyopia and the Pupil
Amblyopia, sometimes called lazy eye, is primarily a brain-based problem where one eye never developed sharp vision during childhood. You might not expect it to affect the pupil, since the optic nerve and retina are structurally normal. Yet a detectable RAPD has been found in a substantial proportion of amblyopic patients. One study found relative afferent pupillary defects in about a third of amblyopic patients tested, with factors like the degree of difference in refractive error between the eyes and the patient’s visual acuity after treatment being the best predictors.11PubMed Central. Pupillary responses in amblyopia Another study examining 55 amblyopic subjects found RAPDs in most of them, with easily visible defects in more than half.12American Journal of Ophthalmology. Pupillary Defects in Amblyopia
Dynamic pupillometry has added detail to this picture. Compared to their healthy fellow eyes, amblyopic eyes showed a longer delay before the pupil started constricting and a shorter overall contraction time, suggesting that the neural processing of the light signal is subtly altered even though the hardware of the eye appears intact.13PubMed Central. Evaluation of Pupillary Light Reflex in Amblyopic Eyes Using Dynamic Pupillometry These defects are typically small, much smaller than what you see with optic neuritis or a retinal vein occlusion, but they are real and reproducible. The clinical significance is that finding a very subtle RAPD in an otherwise healthy-looking child does not always mean optic nerve disease; amblyopia should be on the list.
Using the RAPD to Predict Outcomes
Because the magnitude of a Marcus Gunn pupil reflects how much damage exists along the visual pathway, it can serve as a prognostic tool. This is especially useful after trauma. In patients with traumatic optic neuropathy from facial or orbital injuries, the size of the RAPD at initial evaluation correlates strongly with eventual visual recovery: a larger defect at the start tends to mean worse final vision. One study found a strong negative correlation between RAPD severity and final visual acuity, with a correlation coefficient of −0.83.14PubMed Central. Predictive value of visual evoked potentials, relative afferent pupillary defect, and orbital fractures in patients with traumatic optic neuropathy
The same logic applies in retinal vein occlusion. As noted earlier, the magnitude of the RAPD reliably separates nonischemic cases, which generally have a reasonable prognosis, from ischemic cases, which carry a high risk of complications including abnormal new blood vessel growth.15PubMed. Relative afferent pupillary defect in central retinal vein occlusion In emergency and primary-care settings where advanced retinal imaging may not be immediately available, a simple penlight test can give the clinician meaningful information about how urgently a patient needs to see a specialist.
Brightness comparison tests sometimes supplement the pupil exam. A clinician may ask you to compare the brightness of a light or the redness of an object between your two eyes, scoring each side out of 100. These subjective responses correlate with the RAPD and can add useful information, especially when the pupil exam is difficult to perform, for instance in patients with irregular pupils from previous surgery.16Investigative Ophthalmology & Visual Science. Brightness Sensitivity and Color Perception as Predictors of Relative Afferent Pupillary Defect
Smartphone and Automated Screening
The swinging flashlight test is deceptively simple in concept but tricky to do well in practice. The examiner needs a consistently bright light, a dark room, a cooperative patient, and enough experience to judge whether a pupil truly dilated or just constricted less. These requirements limit who can reliably perform the test, which is a problem when the most important setting for catching a RAPD is often outside the ophthalmology clinic: in the emergency department, the primary-care office, or a rural clinic with no eye specialist on site.
Automated and smartphone-based tools are trying to close this gap. A portable electronic infrared pupillometer can record the pupil’s response during the swinging flashlight test, generating an objective trace that removes the subjectivity of the human observer.17PubMed. Portable pupillography of the swinging flashlight test to detect afferent pupillary defects In glaucoma patients, an automated pupillometer matched an experienced ophthalmologist’s assessment with roughly 89% sensitivity and 92% specificity.18PubMed Central. Quantification of RAPD by an automated pupillometer in asymmetric glaucoma and its correlation with manual pupillary assessment
Even more accessible are smartphone applications. One recently validated cellphone app uses machine learning to locate, segment, and track the pupils during a flashlight test performed by the phone itself, achieving 93% sensitivity and 85% specificity for identifying a RAPD.19PubMed. Validation of a new cell phone application for quick screening of relative afferent pupillary defect A separate pilot study of smartphone infrared pupillography found complete agreement between the phone-based assessment and the real-time clinical exam across all patients tested, including those with anisocoria and reverse RAPDs.20PubMed Central. Pilot Study of Smartphone Infrared Pupillography and Pupillometry These tools are not yet standard of care, but they represent a clear direction. A nurse in a remote clinic who has never performed a swinging flashlight test could, in principle, hold up a phone and get a reliable screening result.
Confusion with Marcus Gunn Jaw-Winking Syndrome
The name “Marcus Gunn” attaches to two completely different clinical findings, which causes endless confusion among students and patients alike. The Marcus Gunn pupil, the subject of this article, is the relative afferent pupillary defect detected by the swinging flashlight test. Marcus Gunn jaw-winking syndrome is a congenital condition in which one eyelid involuntarily lifts when the jaw moves, typically during chewing. The two share nothing except the name of the Scottish ophthalmologist Robert Marcus Gunn, who described both phenomena in the late 19th century. If you have been told you have a “Marcus Gunn” finding, clarifying which one matters enormously: the jaw-winking syndrome is a miswiring of nerve fibers to the eyelid, while the pupillary defect points to a problem with the optic nerve or retina.
When the Test Can Be Misleading
No clinical sign is perfect, and the swinging flashlight test has its blind spots. As the cataract findings illustrate, anything that alters how much light reaches the retina can mimic or mask a true RAPD. Heavily tinted contact lenses, a very small pupil from previous surgery or medication, or simply a brightly lit room can all interfere. Bilateral optic nerve disease, where both sides are equally damaged, produces no relative defect at all, so a normal swinging flashlight test does not rule out serious problems if both eyes are affected.
Examiner technique also matters. The light must be the same brightness and held at the same distance for each eye, and the swing should be brisk. Lingering too long on one side allows the pupil to begin its normal re-dilation cycle, which can be mistaken for a paradoxical dilation. Similarly, the test is harder to interpret in patients with very dark irises, very small pupils, or poor cooperation, which is part of why automated tools are attracting so much interest.
Certain medications complicate things as well. Dilating drops used during an eye exam abolish the pupil’s ability to constrict, making the swinging flashlight test useless until the drops wear off. Pilocarpine and other constricting drops can create an asymmetry that mimics a RAPD. Any interpretation of the test should account for what has been put in the patient’s eyes beforehand.
Even physiologic anisocoria, the mild difference in pupil size that about a fifth of the population has at baseline, can occasionally confuse an inexperienced examiner. True physiologic anisocoria does not change with the swinging flashlight test, but someone unfamiliar with the finding might notice unequal pupils and worry about a neurological problem when none exists. The swinging flashlight test itself resolves this: in physiologic anisocoria, both pupils still constrict equally to the light. The size difference is simply the resting state, not a sign of damage.