Four published studies have looked at RF microneedling around the eyes. All of them are small.
Kim 2023 (n=11). Korean patients aged 30 to 75. Wrinkle scores improved significantly, and biopsies showed denser collagen and elastin in the dermis. But the authors were specific about where the benefit landed: "All patients exhibited wrinkle improvement in the lateral periorbital area, and only two patients also showed efficacy in the lower eyelid area." In other words, crow's feet improved in everyone; the lower lid improved in 2 of 11.
Lin 2025 (n=18). A self-controlled study using needles at 0.5 to 1.5 mm depth, two sessions a month apart, followed for six months with 3D imaging. This one found significant improvement in all four periocular zones — outer corners, lower eyelid, inner corners and upper eyelid. Side effects were mild pain and redness that settled within a week. It is the most encouraging study of the four, and it is still 18 people, each compared only against their own baseline rather than against an untreated group.
Cheles 2024 (n=24). A non-randomised trial, four sessions, three-month follow-up. An aggregated facial wrinkle score fell from 13.00 to 6.09, a mean reduction of 49%. Two of the three authors are employed by the company that makes the device used, and the company co-published the paper. That does not make the result wrong, but it is not independent evidence.
Nilforoushzadeh 2023 (n=9), on dark circles. Skin lightness rose about 32%, measured melanin fell about 49%, and dermal density rose about 30%. Nine people, no control group. This is hypothesis-generating work, not a result you can plan around.
People searching for "under-eye" treatment usually mean one of three things: fine crepey lines under the lash line, a dark shadow, or a puffy or hollow contour.
The published evidence is strongest for the lines at the outer corner. It is weakest for the lower eyelid itself, which is exactly where Kim's 2 of 11 finding sits. And puffiness or hollowing is a volume and anatomy problem, not a collagen problem — no RF microneedling study has been designed to address it.
Worth asking your provider directly which of these they think they can change, and on what evidence.
Lower eyelid skin is roughly 0.5 mm thick — the thinnest on the body. The Lin study used 0.5 to 1.5 mm depths for a reason. Settings that are unremarkable on a cheek or jawline are not appropriate at the orbital rim.
RF microneedling devices can reach several millimetres. Morpheus8's facial tips are labelled to 4 mm and 7 mm depending on the tip. Those depths exist for the jawline and the body, not for an eyelid.
The infraorbital fat compartments — the pads that give the under-eye its contour — sit immediately beneath that very thin skin.
Fat loss is the third most-reported complication in the FDA's MAUDE database for RF microneedling: 26 of 224 reported events between 2013 and 2025, or 11.6% of events. That figure is a share of reports, not a patient risk. MAUDE is voluntary reporting with no denominator, so there is no way to turn it into "X% of people". Under-reporting in cosmetic medicine is heavy.
There is also evidence pointing the other way. A 2024 study measured the submental fat layer by ultrasound after RF microneedling in 30 patients and found no significant change — its title says "without Lipolysis".
Where does that leave the under-eye? No published study has measured periorbital fat volume after RF microneedling. There is no data. What can be said honestly is that this is the area where unwanted volume loss would be the most visible and the hardest to correct, and that the risk is unquantified rather than disproven.
Intraocular corneal shields are standard for energy-device work near the eye. No published periorbital RF microneedling study reports an eye injury — but those studies total roughly 62 patients between them, far too few to pick up a rare event.
No RF microneedling clearance names the eyelid or the periorbital area. Morpheus8's clearances are for electrocoagulation and haemostasis (K192695, 2019) and, since 2023, coagulation or contraction of soft tissue (K231790). Periorbital use with any needle-based RF device is off-label. It is worth being precise about what that does and does not mean: one non-needle radiofrequency device, Thermage's ThermaCool, is cleared for "non-invasive treatment of periorbital wrinkles and rhytids including the upper and lower eyelids" (K052936, 2007) — but that is a different device with a different energy delivery, and its clearance says nothing about needling near the eye. Off-label treatment is legal and common in aesthetics; it simply means the regulator has not reviewed this device for this use, and a clearance is a finding of similarity to an earlier device, not proof that it works.
One more thing readers searching for "Morpheus8" should know: almost none of the published RF microneedling research used Morpheus8. The periorbital studies above used other machines, including Pollogen's Voluderm and a device called MicroRF9.
If you get cold sores, say so when you book, not on the day. Antiviral medicine is started in advance, so it has to be arranged ahead of time. There is no RF microneedling data on cold sore reactivation — the practice is carried over from laser resurfacing, where it is well established.