Cosmetic Laser Eye Protection: What to Expect and Questions to Ask

What proper eye protection looks like at a laser or IPL appointment

If you are booked for laser hair removal, an IPL photofacial, or another cosmetic light treatment, treat eye protection as part of the appointment. For a Class 3B or Class 4 laser, you should be offered eyewear matched to that device. The person firing it, and anyone else who stays in the room, should be wearing protection too. Intense pulsed light is not a laser, but the same practical expectation applies, and at least one state writes IPL into its eyewear rule.

What you should actually have on your face depends on where the handpiece is going.

• Cheeks, jaw, forehead above the brow, neck, and body. Glover and Richer, in a 2023 practical review, describe the usual patient setup as well-fitting periorbital goggles placed over closed eyelids. These are often blackout goggles, meaning visible-light transmission is 0 percent, so you cannot see the room. The same review says the optical density of those cups does not depend on the angle of a stray beam, and that the metal should be brushed or anodized rather than shiny, so the cups themselves do not bounce light back at the operator.

• Eyelids, lash line, or any pass that would require the goggles to move. There is no formal guideline that names a millimeter distance at which goggles must be replaced. Glover and Richer say corneal shields are generally used when treatment is near the eyes, for example on an eyelid with no palpable bone underneath, and whenever goggles would have to be pushed aside to fit the handpiece. A corneal shield is a metal or plastic cover placed on the eye like a contact lens after anesthetic drops. Glover and Richer describe a size that covers a large share of the eye without stretching the lid painfully.

• Operator and staff. They cannot wear the same blackout cups, because they have to see the skin. Glover and Richer report that ANSI guidance recommends at least 20 percent visible-light transmission so staff can still work, and that every pair should be labeled with the wavelength range and the optical density for that range. That 20 percent figure comes from their 2023 account of ANSI guidance, which cited earlier editions. It is not a quotation from the paywalled 2024 text of ANSI Z136.3.

Closing your eyes, wearing sunglasses, or leaving cosmetic contact lenses in are not substitutes. EyeWiki notes that a cosmetic laser can pass a closed lid. Texas's laser rule, which is binding only in Texas, says protective eyewear has to match the wavelength and the optical density the exposure requires, and that the wavelength and optical density must be labeled on the eyewear. In one office IPL injury pooled by Al muqarrab and colleagues, pigment from a cosmetic contact lens was deposited on the cornea and had to be removed surgically. Ordinary contacts do not block the beam.

Published eye injuries are a small literature of case reports and small series, not a count of how often this happens. No one can honestly quote a rate such as "1 in 10,000 treatments," because cosmetic-laser eye injuries are not reported to a single registry. When cases are published, they cluster where protection was absent, taken off to treat near the eye, or matched to the wrong wavelength. Documented outcomes include permanent iris atrophy, uveitis, cataract, secondary glaucoma, and retinal injury. Protection lowers that risk, and injuries are still published in people who had shields or goggles on. In Huang and colleagues' 2018 review of 21 published cases, among the cases in which eye protection was provided, injury still occurred in 33 percent despite corneal shields or wavelength-specific goggles.

Why cosmetic lasers are an eye hazard

The U.S. Food and Drug Administration's laser products page sorts lasers by hazard class. Class IV, the top tier on that table, is an immediate skin and eye hazard from the direct or reflected beam, and the examples include medical lasers used for skin treatment. Class IIIb, the tier below, is an immediate eye hazard when the beam is viewed directly. Clinical reviews usually talk about the same devices as Class 3B and Class 4, which is also the wording in Texas's rule. The FDA page regulates the product: performance standards in 21 CFR 1010 and 1040, and labels that state the class and output. It does not inspect a med spa to see whether anyone put goggles on.

Hair-removal wavelengths sit in the near infrared, past the range the eye sees as a bright aiming beam. Texas commonly describes visible light as 400 to 700 nanometers and treats wavelengths above 710 nanometers as invisible radiation. That covers the usual alexandrite (755 nm), diode (about 800 to 810 nm), and Nd:YAG (1064 nm) hair-removal settings. IPL is different. It is a filtered broad flash, not one wavelength. Al muqarrab and colleagues describe IPL as a broad output from roughly 500 to 1200 nm, so part of the flash is visible and part is not. A visible flash is not a reason to skip eyewear. Laser and IPL are different devices, and the eyewear question still applies to both.

• Alexandrite, 755 nm. EyeWiki describes it as commonly used for hair removal on lighter skin, Fitzpatrick types I to III. It is still a long-wavelength laser in the injury literature. Huang and colleagues say periorbital hair removal, often with alexandrite or diode lasers, is the most reported pattern in facial cosmetic cases.

• Diode, about 800 to 810 nm. EyeWiki describes diode lasers as one of the options with a better skin-safety profile on darker complexions, roughly Fitzpatrick III to V. The same page says longer wavelengths penetrate farther and are common in published ocular injuries. Its tally puts the carbon-dioxide laser first, then alexandrite, diode, and Nd:YAG. A diode session is a reason to be stricter about eyewear. It is not a finding that darker skin makes the eye more fragile.

• Nd:YAG, 1064 nm. EyeWiki describes 1064 nm as the wavelength often favored for hair removal on Fitzpatrick IV to VI skin, because less of it is absorbed in epidermal melanin. Huang's review makes the matching hazard point: high fluence and long wavelength increase ocular injury risk because the energy goes deeper. That is a reason to be stricter about eyewear, not a reason to avoid the wavelength your skin actually needs. Skin-type device choice is a separate decision, covered in the alexandrite, diode, and Nd:YAG comparison and in laser choice for darker skin.

• IPL. Not a laser, and not a single wavelength. Texas still requires protective eyewear for each person exposed to IPL radiation at the levels its rule covers. Al muqarrab's review found that 60 percent of the 20 office eye-injury cases it pooled were IPL, and 40 percent were diode.

Two details from EyeWiki are easy to mix together, and they are not the same claim. First, skin type: longer wavelengths chosen for darker skin are also the ones that show up often in ocular-injury reports, because they penetrate farther. That does not mean darker skin makes the eye itself more fragile. Second, iris color: the same page says lighter irises are more likely to let energy through to the retina, while darker irises more often show iris atrophy, because melanin in the iris absorbs the pulse. A dark-brown iris is not simply "more dangerous" or "safer." It changes which part of the eye is more likely to take the hit if a pulse gets in.

EyeWiki's broader tally of published reports lists the carbon-dioxide laser as the most frequently implicated device, followed by alexandrite, diode, and Nd:YAG. That does not cancel the hair-removal problem. CO2 resurfacing and periorbital hair removal fail in different ways, and both are in the case literature. Near-infrared hair-removal beams are easy to underestimate because you may not see them. Closed lids are not a barrier the reviews are willing to trust.

Reflections matter for Class IV devices. The FDA's Class IV line is an eye hazard from the direct or reflected beam. EyeWiki describes a dermatologist who held a glass slide during an Nd:YAG treatment, reflected the beam into their own eye, and developed foveal choroidal neovascularization two weeks later. Glover and Richer want patient goggles finished matte for the same reason: a polished cup can redirect the beam. Shiny instruments and mirrors are the practical version of that warning. This article does not claim that every reflective surface in a room, from a phone screen to a dental filling, has been measured as a blinding hazard.

graph TD
    A["Class 3B, Class 4, or IPL in use"] --> B{"Protection matched to this device and left in place?"}
    B -->|"No, removed, or wrong wavelength"| C["Closed eyes, sunglasses, contacts, or a hand are not substitutes"]
    C --> D["Reported results include permanent iris atrophy, uveitis, cataract, glaucoma, and retinal injury"]
    B -->|"Goggles, or a corneal shield when treatment is on the lids"| E["Risk is lower. Shielded injuries are still published"]
    E --> F["Huang 2018: among cases in which protection was provided, injury still occurred in 33 percent"]

What published injury reports show

These papers are case reports, a six-patient series, a review of 21 published cases, and one systematic review of 20 office injuries. They show how injuries happened and what the eye looked like afterward. They do not show how often a careful clinic injures someone, because they have no count of uneventful treatments. Read every percentage below as a share of injured people in that paper, not as your personal risk.

Huang also notes a separate litigation review: from 1985 to 2012, laser hair removal was the most commonly litigated cutaneous laser procedure, with four claims involving the eye. Al muqarrab's discussion cites a related litigation series of 174 laser cases, more than half of them hair removal, in which 2.2 percent of the claims were ocular and 0.5 percent cited missing goggles. Those are shares of lawsuits, not shares of treatments. Most of the IPL eye injuries in Al muqarrab's own pool were tied to eyewear that was missing or moved, which is a pattern inside injured patients, not a national frequency.

The 33 percent figure in Huang is the one readers most often misquote. The results section applies it to cases in which eye protection was provided, not to all 21 reports and not to appointments. The useful part is the mechanism attached to it. Protection is not fail-safe when a metal shield heats up, when pulses are stacked without cooling, when a goggle gaps, or when the beam is reflected. EyeWiki adds that plastic shields have melted or caught fire in reports involving Nd:YAG and carbon-dioxide lasers, which is why that page prefers metal shields, and that metal shields can still burn the cornea if they overheat. A shield is a better plan than a lifted goggle. It is not a promise of zero transmission.

Goggles versus corneal shields

External goggles for most facial and body work

For treatment that is not on the eyelids, Glover and Richer describe reusable periorbital goggles over closed lids, often with 0 percent visible-light transmission. Three checks from that review are worth asking to see:

• A label, not a tint. The eyewear should state the optical density and the wavelength or wavelength range that density applies to. Optical density is a log scale. Glover and Richer's worked example is that an optical density of 2 transmits 1 percent of that wavelength. The density you need is the one listed for that device, usually in the operator manual, not a single number that fits every laser. A pair labeled for one wavelength can be the wrong pair for the next handpiece in a multi-wavelength room.

• A matte outside. Stainless-steel patient cups should be brushed or anodized. A polished surface is a mirror.

• A seal that stays put. The cups need to sit around the eye and stay there when you speak or turn. Glover and Richer flag case reports in which protection was removed or pushed aside to make room for the handpiece, and they call that compromise unacceptable.

Corneal shields when the goggles would have to move

Glover and Richer are explicit that no formal guideline sets the distance at which a corneal shield replaces external goggles. In practice they describe shields for treatment near the eyes, such as the upper or lower eyelid where you cannot feel bone under the skin, and for any plan that would otherwise slide the goggles out of the way. Placement is a clinician's procedure: anesthetic drops, a shield sized to cover a large share of the eye without stretching the lid painfully, and a check that the shield is intact so it is less likely to abrade the cornea. Glover and Richer tie that abrasion reduction to sizing and inspection. EyeWiki adds that lubrication can lower the abrasion risk further. The person placing the shield chooses the size, the drops, and the technique.

If the clinic's plan for eyelids, brow hair, or eyeliner-tattoo removal is to lift the goggles or cover the eye with a hand, that matches the failure mode in Huang's 13 of 21 cases and in the Parver hand-shield case. It is reasonable to skip that zone or to wait until someone can explain the shield. The published results include permanent iris damage. They do not say how often any single clinic causes it.

Shields have their own failures. Huang describes injury despite metal corneal shields when the shields overheated during carbon-dioxide resurfacing. EyeWiki notes corneal abrasion as a shield risk, and reflection off metal shields as a reason the operator still needs wavelength-specific goggles. "We use shields, so nothing can reach the eye" is stronger than the case literature.

Questions to ask before treatment

These are questions, not a script you use to supervise the procedure. A clinic that can answer them specifically is easier to judge than one that waves the question off. The clinic still carries the safety duty. Skin burns and pigment change are a different problem from eye injury; they are covered in the laser hair-removal burn guide and the IPL and BBL burn and pigment guide. How hair removal itself works is in the laser hair-removal guide.

Symptoms that mean stop and get an eye exam

Do not try to decide from this page whether an eye is injured. The consistent instruction in the case literature is narrower than a home diagnostic checklist: stop the treatment if the eye hurts or vision changes, and have an ophthalmologist assess the eye promptly. Parver and colleagues put it that way for eyebrow laser injury. Huang's 21 cases were apparent immediately. The common signs were severe eye pain, temporary vision loss, and conjunctival redness. Other series show that "immediate" is not the only clock.

Symptoms reported across these sources, not a ranking of how worried to be:

• Eye pain, including pain that feels inside the eye rather than heat on the forehead skin. In Al muqarrab's pool, ocular pain and visual disturbance were each reported by 60 percent.

• Light sensitivity. Photophobia was the most common symptom in that 20-patient review, at 75 percent.

• Blurred vision, a shadow, a scotoma, or metamorphopsia, which EyeWiki groups as the visual changes patients report. Huang's cases included temporary loss of vision.

• Redness, pupil irregularity, flashing lights, floaters, or swelling. EyeWiki lists these among reported findings. A peaked or poorly reactive pupil is a reason to be examined, not a finding you need to name correctly first.

Timing is part of the warning. In Al muqarrab's 20 patients, symptoms started during the procedure in 7 (35 percent), within a few hours in 4 (20 percent), and within two days in 6 (30 percent). Iris atrophy, when it occurred, was still present months to years later. Separately, the 2024 technician case and EyeWiki's reflected-beam dermatologist case both describe choroidal neovascularization found two weeks after the exposure. Waiting several days to see whether a real visual change fades is not supported. Neither is the idea that every injury is silent for a fortnight. Huang's series was obvious at once.

The next step those reports describe is a prompt examination by an ophthalmologist. Anti-VEGF injections appear only as treatment an ophthalmologist used after neovascularization was diagnosed.

At-home IPL and laser devices

Home devices are a different regulatory object and a thinner evidence file. Al muqarrab and colleagues searched for ocular-injury case reports after home light-based hair removal and did not find published cases of the iris and retinal injuries they were reviewing. They did not find a clean bill of health either. A postmarketing surveillance study they cite logged 34 reports of eye pain after home IPL, of unknown clinical significance. Pain reports are not the same as documented iris atrophy, and they are also not "zero concern."

What the review does show is an overlap in energy, plus a sloppy retail market:

• Office injuries in their pool occurred at 20 to 24 J/cm2. The authors say those fluences sit within home-device output. The home laser they cite as an example, Tria, is described as delivering up to 22 J/cm2. That is not proof that a given home device will injure an eye. It is why they refuse to call the home devices harmless.

• One tested home IPL exceeded a retinal thermal hazard. They cite Eadie and colleagues, who measured a single home IPL and found the retinal thermal hazard exceeded at 11 J/cm2 from 20 centimeters. One device is not the whole market.

• A spot check of 20 devices sold online. Checked against FDA's 510(k) database, 10 of the 20 were not FDA cleared, and 8 of the 20 shipped without goggles. That is a convenience sample of products they happened to find online, not a census of the market.

The authors' own cautions, which match the engineering they describe, are the ones worth keeping. Many of these devices are built so a contact sensor must touch skin before they fire, which is why some manufacturers ship them without goggles and describe them as a low laser class. Leave the contact sensor intact, keep your face out of the aperture, and treat a window aimed at the eye as outside the way these devices are built to fire. Wear the eye protection that came in the box. If none came, do not treat "no goggles in the package" as evidence that eyes are unaffected. Home-device choice, skin-tone limits, and how these compare with clinic hair removal are covered in the at-home laser hair-removal guide. Follow the device instructions, and keep the window off the eyelids and out of an open eye.

What the rules actually require

Patients often assume a single federal rulebook makes every clinic put the same goggles on. The public record is more layered than that, and several of the layers do not govern the appointment at all.

• FDA, the device. The laser products page requires manufacturers to meet 21 CFR 1010 and 1040, including class and output labeling for the higher classes, and it points medical lasers on to device regulation. Class IV on that page includes medical lasers for skin treatment and names both the direct and the reflected beam. The page does not set a clinic's eyewear protocol and does not say FDA inspects whether a patient was handed goggles.

• ANSI Z136.3, voluntary consensus guidance. The Laser Institute's page identifies ANSI Z136.3 (2024), Safe Use of Lasers in Health Care. The public description on that page covers Class 3B and Class 4 lasers used for diagnostic, preventative, cosmetic, or therapeutic care. OSHA's laser-standards page lists the Z136 series as voluntary consensus standards. The 2024 text is sold, so this article does not quote it. The 20 percent visible-light figure for staff stays with Glover and Richer's 2023 review of earlier ANSI guidance.

• OSHA, the employee, not the patient's appointment. OSHA's laser standards page points to general industry rules for personal protective equipment and for eye and face protection, 29 CFR 1910.132 and 1910.133. Those are workplace rules for employees. They are not a cosmetic-patient eyewear code. The page does not say OSHA has a standalone laser standard for med spas.

• Texas, one state's binding rule. Title 25 of the Texas Administrative Code, section 289.301, requires protective eyewear for each individual exposed to laser radiation from IPL, Class 3B, or Class 4 devices. The eyewear must fit around the eye, provide at least the required optical density at the wavelengths in use, be suitable for that wavelength and energy, and carry the optical density and wavelength as a permanent label. It also has to be examined at least every 12 months. Class 4 controlled-area signs must say that laser eye protection is required and must state optical density, laser type, wavelength, and wattage. That section binds Texas registrants. Other states write different rules. A federal device label does not, by itself, show that this eyewear rule has been met.

• Reporting. Glover and Richer describe a lack of enforced regulation around these devices and no mandated reporting of adverse events, which keeps the true ocular-injury incidence unknown. Manufacturer reports to FDA are a different system and do not turn every in-office eyewear lapse into a public case. Absence of a rate is not evidence that injuries do not happen. It is why this page refuses to invent one.

Clinic operators building a laser safety program, naming a laser safety officer, and keeping eyewear logs are doing a different job from the one this article is for. That operator guide is separate. Your questions do not replace that program. They tell you whether the room you are sitting in can describe the controls the case literature actually talks about.

Where the evidence stops

• No incidence rate. Every percentage in the injury papers is a share of published injuries, of the protected cases inside those reports, or of lawsuits. Uneventful treatments were not counted. Do not convert 13 of 21, Huang's 33 percent of protected cases, or 5 of 40 into the chance that your appointment goes wrong.

• Protection reduces injury in these reports and does not eliminate it. Huang's 33 percent, Parver's shielded patients, overheated metal shields, and melted plastic shields are the counterexamples. "Zero risk once the cups are on" is not what those papers found.

• Home-device harm is mostly inferred. The 2023 systematic review found no published structural eye-injury case reports from the home devices themselves. It did find overlapping fluences, a single-device hazard measurement, a 20-product spot check with missing clearances and missing goggles, and 34 eye-pain reports of unknown significance. That is a warning, not a case series of blinded consumers.

• ANSI Z136.3 (2024) was not quoted clause by clause. The edition and its public scope are verifiable. The paid text is not, so staff visible-light transmission and similar figures stay attributed to the 2023 review.

• Texas is not the country. Section 289.301 is a useful example of a binding eyewear rule, including IPL. It does not travel with you across a state line.

• Skin type is not an eye-color rule. Longer wavelengths often used on darker skin are also deep-penetrating ocular hazards. Lighter and darker irises tend to show different injury patterns if a pulse gets in. Neither statement is a personal risk score, and neither is advice to pick a wavelength from this page.

Use this as a reading of public regulator pages, one state rule, and the case literature above, so you can ask better questions at the appointment. Diagnosis stays with an ophthalmologist, and the clinic remains responsible for the setup in the room.

Sources

• U.S. Food and Drug Administration. Laser Products and Instruments. Hazard classes, including Class IV medical lasers for skin treatment as an immediate skin and eye hazard from the direct or reflected beam, and the performance standards in 21 CFR Parts 1010 and 1040.

• Parver DL, Dreher RJ, Kohanim S, et al. Ocular Injury After Laser Hair Reduction Treatment to the Eyebrow. Archives of Ophthalmology. 2012;130(10):1330-1334. Six-patient series of eyebrow laser injuries, including hand shielding and permanent iris damage.

• Huang A, Phillips A, Adar T, Hui A. Ocular Injury in Cosmetic Laser Treatments of the Face. Journal of Clinical and Aesthetic Dermatology. 2018;11(2):15-18. Review of 21 case reports: protection absent or removed in 13, and injury also reported despite shields or wavelength-specific goggles.

• Glover C, Richer V. Preventing Eye Injuries From Light and Laser-Based Dermatologic Procedures: A Practical Review. Journal of Cutaneous Medicine and Surgery. 2023;27(5):509-515. Patient goggles, corneal shields, labeled optical density, staff visible-light transmission, and the lack of a reported incidence rate.

• Al muqarrab FJ, Alsuwaidan S, Al muqarrab Z, Alajlan A. Is It Possible for Light-Based Hair Removal Home Devices to Induce Ocular Damage? Systematic Review. Clinical, Cosmetic and Investigational Dermatology. 2023;16:3731-3742. Twenty office injuries, home-device fluence overlap, and a 20-device online spot check.

• Mohammad NK, et al. Severe Retinal Damage and Choroidal Neovascularization Following Accidental Laser Exposure During Hair Removal: A Case Report. Cureus. 2024;16(9):e69075. Technician injured by a CLARITY II discharge without protective glasses; subfoveal neovascularization at two weeks.

• American Academy of Ophthalmology. Ocular Injuries Due to Cosmetic Laser. EyeWiki. Professionally edited clinical summary, not a peer-reviewed trial. Wavelength patterns, eyewear failures, corneal shields, and iris-color observations. Page review dated September 18, 2026.

• Laser Institute of America. ANSI Z136.3, Safe Use of Lasers in Health Care. Public description of the 2024 health-care laser standard, including Class 3B and Class 4 lasers and cosmetic use. The standard text is not quoted here because it is not freely available.

• Occupational Safety and Health Administration. Laser Hazards: Standards. Points to general eye and face protection rules for workers and lists ANSI Z136 as voluntary consensus guidance, not an OSHA laser regulation.

• Texas Department of State Health Services. 25 TAC § 289.301, Radiation Requirements for Lasers and IPL Devices. Binding Texas requirements for eyewear fit, wavelength, optical density, labeling, and Class 4 signage. Not a national mandate.