Bottom line
This distinction matters. The relevant scientific question is not whether the FDA has cleared the FX-635 for a brain indication—it has not—but whether low-intensity red light in the same spectral neighborhood has been studied transcranially in humans, whether plausible biological mechanisms exist, and what the present evidence says about safety, dosing, penetration, and uncertainty.
What the FX-635 actually is
Human transcranial PBM evidence relevant to red light near 635 nm
1. Chronic traumatic brain injury: 633 nm + 870 nm
Naeser and colleagues reported transcranial treatment using LED cluster heads containing 633 nm red and 870 nm near-infrared diodes in people with chronic traumatic brain injury. Early case reports described improvements in sustained attention, executive function, and memory. These studies were preliminary and small, but they are directly relevant to the question of whether visible red light around 633 nm has been intentionally applied to the human head in published neurologic research.
PubMed: Improved Cognitive Function After Transcranial LED Treatments in Chronic TBI
2. Older adults: randomized sham-controlled trial using 633 nm + 870 nm
A randomized, sham-controlled study of 30 older adults applied three LED clusters containing 633 nm and 870 nm light to frontal and posterior scalp locations for 7.5 minutes. The active PBM group showed significant improvement in measures of action selection, inhibition, and mental flexibility compared with sham. The study does not isolate the effect of 633 nm from 870 nm, but it demonstrates deliberate transcranial use of 633 nm red light in a controlled human experiment.
Full text: Photobiomodulation Improves the Frontal Cognitive Function of Older Adults
3. Parkinson's disease: randomized sham-controlled 635 nm + 810 nm helmet
A purpose-built transcranial helmet containing 20 red 635 nm LEDs and 20 near-infrared 810 nm LEDs was evaluated in a blinded randomized trial in Parkinson's disease. Treatment was generally well tolerated. The total motor score did not show a significant active-versus-sham difference, although some motor subdomains improved and later analyses reported responder effects. This is particularly relevant because the visible red component was exactly 635 nm.
Full text: Randomized Placebo-Controlled Study of a Transcranial PBM Helmet in Parkinson's Disease
Full text: Randomized feasibility study of 635 nm + 810 nm tPBM in Parkinson's disease
4. Chronic stroke and aphasia: 633 nm + 870 nm with fMRI outcomes
A small chronic-stroke case series used 633 nm + 870 nm transcranial LED clusters and assessed language outcomes together with functional MRI. Some treatment protocols were associated with improved naming and altered functional connectivity. This remains low-level evidence because of the small sample and case-series design, but it adds another neurologic population in which 633 nm red light was intentionally delivered transcranially.
PubMed: Red/NIR transcranial PBM in chronic stroke with aphasia
5. Systematic reviews of human tPBM
A 2022 systematic review of human cognition studies identified 35 studies; 29 reported improvement in at least one cognitive outcome. The authors emphasized that only about half of the clinical trials were randomized controlled trials and that protocols varied substantially. Importantly for wavelength comparison, the review states that human protocols ranged from visible red light at approximately 630–635 nm through near-infrared wavelengths above 1,000 nm.
A separate systematic review evaluated changes in human brain activity associated with tPBM and concluded that the field has meaningful evidence of physiologic brain effects, while emphasizing heterogeneity and the need for more rigorous trials.
PubMed: Systematic review of tPBM effects on brain activity in humans
Why red / near-infrared light is biologically plausible
Photobiomodulation is fundamentally different from ablative or heating laser procedures. PBM uses relatively low optical power to initiate photochemical and cell-signaling effects rather than relying on tissue heating.
A frequently proposed mechanism involves mitochondrial chromophores, particularly cytochrome-c oxidase (Complex IV), with downstream changes in mitochondrial membrane potential, oxygen utilization, ATP production, nitric-oxide signaling, reactive oxygen species signaling, transcription factors, inflammatory pathways, and cellular stress responses. Neural tissue is metabolically demanding and mitochondria-rich, which is one reason brain PBM has become an active research area.
The mechanistic literature is not equivalent to proof that a specific clinical protocol treats a specific neurologic disorder. It does, however, provide a biologically coherent rationale for investigating low-intensity red and near-infrared light in neural tissue.
Full text: Brain Photobiomodulation Therapy — A Narrative Review
Full text: Photobiomodulation and the Brain — A New Paradigm
An important limitation: wavelength and penetration
Red light near 633–660 nm does not penetrate the adult human scalp and skull as efficiently as wavelengths in the near-infrared “optical window,” particularly around 800–850 nm. Cadaver and modeling studies generally show substantially better transcranial transmission for near-infrared wavelengths than for 633 nm red light.
This matters when interpreting the FX-635. The fact that 633–635 nm has been included in human transcranial studies does not mean that a 635–640 nm source produces the same cortical dose as an 810–830 nm device. Treatment geometry, contact versus non-contact delivery, beam area, irradiance, exposure time, scalp/hair characteristics, skull thickness, wavelength, pulsing, and total fluence all influence how much light can reach deeper tissue.
Full text review: Shining Light on the Head — Photobiomodulation for Brain Disorders
Safety and clinical caution
Across published tPBM studies, serious treatment-related adverse events have been uncommon. Reviews describe the technique as generally well tolerated, while also noting that study sizes are often small, protocols are heterogeneous, and long-term safety data for repeated treatment remain less developed than for established medical therapies.
Reported minor effects in various tPBM studies have included headache, transient dizziness, fatigue, sleep disturbance, irritability, and localized warmth. These findings come from different wavelengths and devices and cannot automatically be transferred to every laser system or dose.
Eye safety remains important. Laser-class precautions and manufacturer safety requirements should be followed. Direct ocular exposure is not part of the rationale for transcranial PBM.
Full text: Tolerability and Safety of Transcranial Photobiomodulation
PubMed: Dose-dependent tolerability and safety of transcranial PBM — randomized controlled trial
How I would characterize use of the FX-635 over the head
A scientifically conservative description would be:
“The FX-635 is an FDA-cleared, non-thermal low-level red laser used clinically for musculoskeletal pain. Its FDA clearance does not include treatment of the brain or neurologic disease. Separately, transcranial photobiomodulation is an established research field, and published human studies have intentionally applied red light at approximately 633–635 nm to the scalp, commonly alongside near-infrared light, while investigating cognitive and neurologic outcomes. Cranial use of the FX-635 should therefore be understood as an off-label extrapolation from the broader PBM literature—not as an FDA-cleared neurologic treatment and not as a device-specific claim of proven efficacy.”
This wording preserves the relevant science without claiming that evidence generated with a different device, dose, or wavelength combination is proof of a clinical effect from the FX-635 itself.
Selected references
Primary source for the FX-635 regulatory indication and non-thermal/non-heating device classification.
FDA recordTranscranial LED clusters included 633 nm red and 870 nm near-infrared diodes.
PubMedReview and clinical research program involving red/NIR transcranial PBM in chronic TBI.
Journal pageRandomized sham-controlled human study using LED clusters containing 633 nm and 870 nm light.
Full textTriple-blind randomized trial using a helmet containing 635 nm red and 810 nm near-infrared LEDs.
Full textRandomized, sham-controlled feasibility trial of 635 nm + 810 nm transcranial PBM.
Full textSmall chronic-stroke case series using 633 nm + 870 nm LED clusters with language and fMRI outcomes.
PubMedSystematic review of 35 human studies; wavelengths in reviewed protocols ranged from visible red at 630–635 nm to NIR above 1,000 nm.
PubMedSystematic review focusing on human brain-activity outcomes.
PubMedSystematic review addressing efficacy, treatment parameters, and safety in brain disorders.
PubMedDetailed review of proposed mitochondrial, inflammatory, neuroprotective, and neuroplastic mechanisms.
Full textMechanistic and translational review of red/NIR PBM applied to brain tissue.
Full textReview including transcranial penetration data and the wavelength-dependent limitation of visible red light.
Full textReview and analysis focused on tolerability and adverse effects of repeated tPBM.
Full textRandomized sham-controlled study in healthy adults comparing 660 nm and 850 nm continuous and pulsed tPBM.
PubMedClinical and regulatory note
This page is an educational review of published research and device information. It is not a claim that the Erchonia FX-635 is FDA-cleared to diagnose, prevent, or treat traumatic brain injury, dementia, Parkinson's disease, stroke, depression, cognitive impairment, or any other neurologic condition. Published studies cited here used a variety of devices, wavelengths, optical powers, beam geometries, treatment locations, and doses. Their results should not be assumed to establish efficacy or dose equivalence for the FX-635. Clinical decisions should be individualized and made within applicable professional scope, device-safety requirements, informed-consent standards, and governing law.
Research summary prepared for professional discussion. Last updated August 20, 2026.