Skip to Content

Central Retinal Artery Occlusion (CRAO): Sudden Vision Loss & Eye Stroke Guide (2026 Guide)

A blocked retinal artery is not merely an eye problem — it is the same ischemic event as a brain stroke, just visible through an ophthalmoscope. The roughly 90-minute ischemic tolerance ceiling, why the cherry red spot appears, the giant cell arteritis that must never be missed, and why every patient needs a same-day vascular work-up regardless of what happens to their vision.
Central Retinal Artery Occlusion (CRAO): Sudden Vision Loss & Eye Stroke Guide | Agaaz Ophthalmics

Retinal Vascular Emergencies · Surgeon Series

Sudden. Painless.
One eye. This is
a stroke of the retina.

Central retinal artery occlusion blocks the eye's own blood supply the same way a cerebral stroke blocks the brain's — except it happens in an organ you can look straight into with an ophthalmoscope. The retina tolerates interrupted blood flow for minutes, not hours, and the outcome turns on how fast the blockage is recognised and where the underlying clot actually came from.

~90 minanimal-model ischemic
tolerance ceiling
2021AHA/ASA classified CRAO
as a stroke subtype
0FDA-approved drug
treatments for CRAO
12 minreading time

Section 01 — The Eye's Own Stroke

Blocked artery.
Starved retina.
The clock starts immediately.

Patients describe it almost identically: vision in one eye simply went, with no pain, no warning, often on waking or mid-conversation. That description is not incidental — it is the signature of an embolic event, the same signature a stroke neurologist would recognise instantly if it happened in a limb or in speech instead of in an eye.

Central retinal artery occlusion (CRAO) is a sudden blockage of the central retinal artery — the vessel, branching from the ophthalmic artery, that supplies blood to the inner layers of the retina. Most cases are embolic: a fragment of material, usually cholesterol plaque from the carotid arteries or a clot from the heart, travels downstream and lodges in a vessel too narrow to pass. The retina has one of the lowest tolerances for interrupted blood flow of any tissue examined, which is why this condition is managed as a genuine emergency even though, honestly, the treatment options available once the blockage has happened are limited.

The core answer, in 130 words

Central retinal artery occlusion (CRAO) is a sudden blockage of the eye's main blood supply, almost always by an embolus travelling from the carotid arteries or the heart, and it behaves exactly like an ischemic stroke — because that is what it is, just downstream in the eye rather than the brain. Vision loss is typically sudden, painless and profound in the affected eye. Unlike a limb, the retina has almost no tolerance for interrupted blood flow: animal-model data suggests meaningful, irreversible damage begins within roughly 90 minutes, though the exact human threshold is not known with precision. Because the embolic source causing a CRAO is the same one that causes cerebral stroke, current guidance treats every CRAO patient as needing the same urgent vascular work-up as a stroke or TIA, regardless of whether useful vision ever returns.

Mechanism
A retinal artery — usually the central retinal artery itself, sometimes just a branch — is blocked, most often by an embolus, cutting off the inner retina's only blood supply.
What's affected
CRAO typically wipes out most or all of the visual field in that eye; a branch occlusion (BRAO) affects only the sector of retina fed by that branch.
Reversibility
Unlike a blocked outflow vein, arterial blockage leaves almost no time margin — the retina has one of the lowest ischemic tolerances of any tissue in the body.
Systemic stakes
The clot did not originate in the eye. Finding where it came from — carotid, heart, or elsewhere — is arguably more urgent than anything done to the eye itself.

Section 02 — Interactive

The number that decides
everything: minutes.

Move the clock forward and watch how an ischemic retina is presumed to change at each stage, based on the animal-model tolerance data underlying current emergency guidance. This is a simplified, illustrative model of tissue-level change over time — not a diagnostic tool, and not a precise human timeline, since human retinal tolerance has never been measured this exactly.

Interactive: retinal ischemia vs. time since onset
The retina's tolerance for interrupted blood flow is estimated from primate experiments, not measured directly in humans. This model illustrates the ischemic ceiling those experiments established — useful for the 'time is retina' emergency message, not as a precise human countdown.
0 min
drag to rotate
Retina Embolus / blockage Cherry-red spot (fovea)

Within the first hour or so, the retina is under severe metabolic stress but — in principle — still salvageable if flow is restored; this is the rationale behind bedside measures like ocular massage or anterior chamber paracentesis, though none has been reliably shown in controlled trials to restore flow. Past roughly 90 to 100 minutes, Hayreh's primate experiments found retinal damage became substantial, and by around four hours, close to complete. Human eyes have not been subjected to controlled experiments of this kind, for obvious reasons, and case reports describe partial recovery after longer intervals than the animal data would predict — so the 90-minute figure functions as a conservative biological ceiling used to justify treating CRAO as a true emergency, not as a stopwatch that determines whether treatment is worth attempting.

Why the number is imprecise, and why guidance uses it anyway. No ethics board is going to time exactly how long a human retina can go without blood flow. The roughly 90-minute figure comes from primate experiments (Hayreh and colleagues) measuring the point at which retinal damage became histologically substantial. Human retinal tolerance likely varies with collateral circulation, the presence of a cilioretinal artery, and how completely the vessel is blocked — but the number persists in guidelines because it carries the right message: whatever the true human ceiling is, it is measured in minutes to a few hours, not in the days it might otherwise take a patient to reach an ophthalmologist.

Section 03 — Why Speed Decides Everything

Same blockage,
different address.
The prognosis moves with it.

Not every retinal artery occlusion behaves the same way. Where exactly the blockage sits — the ophthalmic artery, the central retinal artery trunk, or a single branch — changes both what the eye exam shows and how much vision has a realistic chance of surviving.

The 'cherry red spot' is the single most recognisable sign of CRAO on fundus exam. As the inner retina infarcts, the nerve fibre and ganglion cell layers swell and turn opaque white, most visibly around the macula where those layers are thickest. The fovea itself has almost none of that inner retinal tissue — light reaches the photoreceptors almost directly there — and sits over the still-intact choroidal circulation underneath. Against the surrounding pallor, the thin, choroid-perfused fovea shows through as a comparatively red spot at the centre of a pale retina. It is a contrast artefact of anatomy, not evidence that central vision has been spared.

Occlusion siteField defectCherry-red spot?General prognosis
Ophthalmic artery occlusionSevere, often near-total field lossOften absent — choroidal supply is also compromisedGenerally the worst of the group
Central retinal artery occlusion (trunk)Near-total field loss, aside from any cilioretinal-sparing islandClassically presentPoor without spontaneous reperfusion or a sparing vessel
Branch retinal artery occlusion (BRAO)Sectoral / altitudinal defect matching the branch's territoryLocalised pallor only, not a full spotGenerally better — unaffected retina is preserved
Transient monocular vision loss (amaurosis fugax)Curtain-like, fully reversible within minutesNot applicable — no infarction if truly transientVision recovers, but carries the same warning value

A cilioretinal artery — a small vessel arising directly from the choroidal circulation rather than the central retinal artery — is present in a meaningful minority of eyes and, when it supplies part of the macula, can preserve a central island of vision even through an otherwise complete CRAO. It is one of the few pieces of genuinely good anatomical luck in this condition, and one reason funduscopic findings are examined carefully rather than assumed from the diagnosis alone.

Section 04 — Where the Embolus Comes From

The clot did not
start in the eye.
That is the real emergency.

Treating CRAO as purely an ophthalmic event misses the point. In the large majority of cases, the retinal artery is a downstream casualty of disease sitting somewhere else entirely — the carotid arteries, the heart, or, in a subset of older patients, an inflamed artery wall that has nothing to do with cholesterol at all.

Embolic — carotid
Cholesterol emboli shed from atherosclerotic plaque in the carotid arteries, visible on exam as glistening Hollenhorst plaques, are the most commonly identified embolic source.
Embolic — cardiac
Atrial fibrillation, valvular disease and other cardiac sources can throw platelet-fibrin or calcific emboli into the ophthalmic circulation exactly as they do into the cerebral circulation.
Arteritic — giant cell arteritis
In patients over roughly 50, giant cell arteritis must be actively ruled out. Missed arteritic CRAO can blind the second eye within days and carries its own risk of aortic and other large-vessel complications.
Non-embolic / other
Hypercoagulable states, vasospasm, orbital compression and, rarely, trauma or drug-related causes account for a smaller share, more often in younger patients without the usual atherosclerotic risk factors.

Giant cell arteritis deserves special mention because it is the one cause of CRAO where speed of a different kind matters: speed of steroid treatment, not speed of ophthalmic intervention. Jaw claudication, scalp tenderness, new temporal headache, unexplained weight loss and polymyalgia symptoms in a patient over 50 presenting with sudden vision loss should prompt same-day ESR and CRP testing and, in most protocols, empiric high-dose systemic corticosteroids started before a confirmatory temporal artery biopsy — because waiting for biopsy confirmation risks the fellow eye.

Section 05 — What the Evidence Actually Shows

The eye's stroke,
on the same
guidelines as the brain's.

The clinical literature on CRAO over the past two decades has moved in one consistent direction: away from treating it as a self-contained eye problem, and toward treating it as a cerebrovascular event that happens to present to an ophthalmologist first.

In 2021, the American Heart Association and American Stroke Association issued a scientific statement formally classifying acute retinal ischemia, including CRAO, as a stroke subtype, and recommending the same urgent evaluation pathway used for cerebral TIA and stroke — same-day imaging, carotid and cardiac evaluation, and secondary-prevention treatment, regardless of whether the eye itself ever recovers vision. That recommendation did not appear from nowhere: diffusion-weighted MRI studies in CRAO patients have repeatedly found a meaningful minority with an acute, clinically silent brain infarct elsewhere at the time they present with eye symptoms — vision loss the patient noticed, brain injury they did not.

On the treatment side, the evidence is far less reassuring than patients usually hope. Traditional bedside measures — ocular massage, anterior chamber paracentesis to lower intraocular pressure, breathing a carbon-dioxide-enriched gas mixture to dilate retinal vessels, hyperbaric oxygen — have circulated in the literature for decades without controlled-trial evidence that any of them reliably restores vision, a gap summarised bluntly in a widely cited 2018 editorial urging clinicians to follow evidence-based emergency guidelines rather than reflexive bedside manoeuvres. Intravenous thrombolysis, the definitive treatment for many ischemic cerebral strokes, was tested specifically for CRAO in the European EAGLE trial; the trial did not demonstrate a visual benefit over conservative management sufficient to justify the added bleeding risk, and it remains the strongest controlled evidence against routine thrombolysis for CRAO outside a research protocol. At the time of writing, there is no FDA-approved pharmacologic treatment specific to CRAO.

AHA/ASA classification of CRAO as a stroke subtypeFormal consensus
CRAO patients found to have a concurrent silent brain infarct on MRIA meaningful minority
Visual benefit of IV thrombolysis beyond conservative care (EAGLE trial)Not demonstrated
Traditional bedside measures shown to restore vision in controlled trialsNo controlled evidence

Illustrative visualisation of the direction and rough weight of the cited literature — see References for source studies. Not pooled meta-analytic effect sizes or precise percentages.

0approx. ischemic tolerance
ceiling, animal-model estimate
0age threshold prompting
mandatory GCA rule-out
0FDA-approved drug treatments
specific to CRAO

Section 06 — Real Risk, Honestly Stated

Doing nothing,
and doing something,
both carry risk.

It would be dishonest to describe CRAO management as a simple 'try everything, it can't hurt' emergency. Some traditional bedside interventions carry their own risk, thrombolysis carries a real bleeding risk, and the single biggest risk of all is treating this as an eye-only problem and skipping the vascular work-up entirely.

RiskWhy it matters in CRAO
Missed giant cell arteritisDelayed steroid treatment risks bilateral blindness within days and systemic vascular complications — the single most time-critical diagnosis to exclude
Recurrent stroke or heart attack without work-upThe embolic source causing the CRAO remains untreated and can cause a subsequent, potentially disabling cerebral or cardiac event
Thrombolysis bleeding riskIV tPA carries a real intracranial and systemic haemorrhage risk that, per the EAGLE trial, was not clearly offset by visual benefit
Anterior chamber paracentesisAn invasive bedside procedure with its own small risk of infection, hyphema or lens injury, for a benefit that remains unproven
Neovascular complicationsLess common than after an ischemic central retinal vein occlusion, but iris and angle neovascularisation are documented after CRAO and warrant monitoring for months afterward
Permanent field loss despite guideline-concordant careEven prompt, correct management does not reliably restore central vision once infarction has occurred — the retina's poor collateral supply is the fundamental limit, not a failure of treatment
The honest framing. Ophthalmology does not have a CRAO treatment with the reliability of a coronary stent restoring cardiac blood flow. What it does have is a well-supported argument that CRAO is a systemic vascular event, and that finding and treating the source — carotid disease, atrial fibrillation, giant cell arteritis — protects the rest of the patient's vascular territory even when the affected eye's vision does not come back. That is a less satisfying message than a rescue procedure, and it is the one the evidence actually supports.

Section 07 — Where the Decision Actually Gets Made

The eye exam ends.
The stroke
work-up begins.

In a well-run pathway, the ophthalmologist's job is not just to diagnose CRAO on fundus exam — it is to trigger the same urgent referral a neurologist would trigger for a TIA, on the same day, before the patient leaves the building.

That work-up looks functionally identical to a TIA clinic pathway, run in parallel with the ophthalmic assessment rather than instead of it: same-day carotid duplex ultrasound, an electrocardiogram and often extended cardiac monitoring for paroxysmal atrial fibrillation, echocardiography where a cardioembolic source is suspected, same-day ESR and CRP with a low threshold for temporal artery biopsy in patients over 50, and a full lipid and vascular risk-factor review.

Illustrative Scenario — Composite, Not an Individual Patient
The eye that found the carotid disease

A patient in their late sixties wakes up with painless, complete loss of vision in one eye and is seen the same morning. Fundus exam shows retinal whitening and a cherry-red spot consistent with CRAO. No headache, no jaw pain, nothing pointing toward arteritis. What the eye exam cannot show is why the artery blocked — that question sends the patient for carotid duplex imaging the same day, which finds a high-grade stenosis on the side matching the affected eye.

The vision in that eye does not meaningfully recover; by the time of presentation, the ischemic window described in the animal-model data had already closed. But the carotid finding changes the rest of the patient's care — evaluation for endarterectomy or stenting, antiplatelet therapy, risk-factor management — precisely the intervention that reduces the risk of a subsequent, and potentially far more disabling, cerebral stroke. The eye did not save its own vision. It found the disease that could have taken far more.

Illustrative composite consistent with the AHA/ASA 2021 scientific statement's recommended work-up pathway for acute retinal ischemia — not a specific patient record.

Section 08 — FAQ

Frequently asked questions
about CRAO.

CRAO is a sudden blockage of the central retinal artery, the vessel that supplies blood to the inner layers of the retina. It typically causes sudden, painless, and often severe loss of vision in one eye, and is usually caused by an embolus — a fragment of clot or cholesterol plaque — travelling from the carotid arteries or the heart.

Functionally, yes. The retina is technically part of the central nervous system, and a CRAO is caused by the same kind of vascular blockage — usually an embolus — that causes an ischemic stroke in the brain. In 2021 the American Heart Association and American Stroke Association formally classified acute retinal ischemia as a stroke subtype and recommended the same urgent vascular work-up used for cerebral stroke or TIA, regardless of whether vision in the eye recovers.

It's the classic fundus exam finding in CRAO. The retina around the macula turns pale as the inner retinal layers infarct and swell, while the very centre of the macula — the fovea, which has almost no inner retinal tissue and sits over the still-intact choroidal blood supply — stays comparatively red by contrast. It's a sign of the pattern of damage, not evidence that central vision has been spared.

Sometimes, but it is unpredictable and often limited. Recovery depends heavily on how quickly blood flow is restored, whether a protective cilioretinal artery happens to spare part of the macula, and how complete the original blockage was. Animal-model data suggests the retina's tolerance for complete ischemia is on the order of 90 minutes to a few hours before damage becomes substantial, which is why CRAO is treated as a true time-critical emergency even though, honestly, no treatment reliably reverses established damage.

The eye exam confirms the diagnosis, but the more important work happens afterward: same-day carotid duplex ultrasound, an ECG and often extended heart-rhythm monitoring, blood tests including ESR and CRP to screen for giant cell arteritis, and a broader cardiovascular risk-factor review. The goal is finding the embolic or inflammatory source before it causes a second, potentially more disabling event — commonly a cerebral stroke.

Giant cell arteritis is an inflammatory disease of medium and large arteries that, in patients typically over 50, can cause an arteritic form of CRAO. It matters because it is treated completely differently from embolic CRAO — with urgent high-dose systemic corticosteroids, often started before a confirmatory temporal artery biopsy — and because an untreated arteritic CRAO carries a real risk of blinding the second eye within days if not recognised.

References & Evidence Base

Peer-reviewed
citations.

Mac Grory B, Schrag M, Biousse V, et al. "Management of Central Retinal Artery Occlusion: A Scientific Statement From the American Heart Association." Stroke. 2021. PubMed search
Varma DD, Cugati S, Lee AW, Chen CS. "A review of central retinal artery occlusion: clinical presentation and management." Eye (Lond). 2013. PubMed search
Hayreh SS, Zimmerman MB, Kimura A, Sanon A. "Central retinal artery occlusion. Retinal survival time." Exp Eye Res. Foundational and follow-up primate ischemia-tolerance work. PubMed search
Schrag M, et al. (EAGLE trial investigators). Randomized European trial of intravenous thrombolysis versus conservative management for central retinal artery occlusion. PubMed search
Biousse V, Nahab F, Newman NJ. "Management of Acute Retinal Ischemia: Follow the Guidelines!" Ophthalmology. 2018. PubMed search
Lauda F, et al. Diffusion-weighted MRI studies of concurrent acute silent brain infarction in patients presenting with monocular ischemic vision loss. PubMed search
American Academy of Ophthalmology. "Central Retinal Artery Occlusion" — EyeWiki clinical reference on presentation, work-up and management. eyewiki.aao.org
StatPearls / NCBI Bookshelf. "Central Retinal Artery Occlusion" clinical reference chapter. NCBI Bookshelf

Continue Reading

Related guides
from Beyond Vision.

Start writing here...