How to Test Facial Nerve Function: Methods and Results

Testing facial nerve function involves a combination of bedside clinical examination, standardized grading scales, and electrodiagnostic tools, each suited to different stages and types of facial nerve problems. The simplest tests require nothing more than asking a patient to raise their eyebrows, close their eyes tightly, and smile, while the most sophisticated use electrical stimulation and needle electrodes to measure nerve conduction at a cellular level. Which test matters most depends on what you are trying to find out: the severity of weakness right now, whether the nerve is degenerating, or how likely someone is to recover.

Bedside Clinical Examination

The foundation of facial nerve testing is a hands-on exam. A clinician watches you perform a series of facial movements, comparing the two sides of the face. The key movements are raising the eyebrows (testing the forehead branch), closing the eyes tightly (the zygomatic and temporal branches), puffing out the cheeks, showing the teeth, and pursing the lips (the buccal and marginal mandibular branches). Each of these targets a different branch of the facial nerve, so weakness in one zone but not another helps localize where the problem lies.

The distinction between the upper and lower face matters for a critical reason. If only the lower half of the face is weak but the forehead still works, the problem is almost certainly in the brain rather than the nerve itself. That is because the forehead muscles receive signals from both hemispheres of the brain, giving them a backup supply. When the peripheral nerve itself is damaged, both upper and lower zones go weak on the same side. This single observation often determines whether imaging of the brain is urgently needed.

Beyond motor function, the facial nerve also carries taste fibers from the front two-thirds of the tongue, controls tear production through the lacrimal gland, and regulates a tiny muscle in the middle ear called the stapedius. Tests for these parasympathetic and sensory branches help pin down the exact location of a lesion along the nerve’s winding path from the brainstem through the skull. The Schirmer test, which measures tear production by placing a small paper strip under the lower eyelid, and electrogustometry, which applies a weak electrical current to the tongue to test taste thresholds, are among the classic tools for this purpose.¹ Taste testing after ear surgery, for example, can confirm whether the chorda tympani nerve, a branch of the facial nerve running through the middle ear, has been injured.²

Clinical Grading Scales

Once weakness is identified, clinicians need a way to quantify it. The most widely used tool is the House-Brackmann (HB) scale, a six-point system ranging from Grade I (normal function) to Grade VI (complete paralysis). It is quick, requires no equipment, and has been the global standard since the 1980s. Its weakness, though, is that it treats the entire face as a single unit and relies heavily on subjective impression. Studies consistently find only moderate agreement between different raters using the HB scale, with one study reporting an overall inter-reader kappa of about 0.54 for global scores, with even lower agreement for individual zones like the eye and forehead.³ Another study found that while the HB system had substantial reliability overall, the exact percentage agreement between trained observers was only 44%, with wide variation on individual grades.⁴

To address these shortcomings, newer scales have been developed. The Sunnybrook Facial Grading System and the Facial Nerve Grading System 2.0 (FNGS 2.0) both break the face into separate regions and assign scores to each, capturing regional differences that the HB scale misses. Both also account for synkinesis, the involuntary twitching in one facial zone when you deliberately move another. A direct comparison of all three systems found that the FNGS 2.0 and Sunnybrook systems achieved excellent inter-rater reliability, with intraclass correlation coefficients above 0.95, while the HB scale showed only moderate agreement with kappa values around 0.4 to 0.5.⁵

Despite its limitations, the HB scale remains entrenched in clinical practice and research reporting. Clinician experience plays a role in how well it performs; agreement tends to increase among more experienced raters.³ In many settings, clinicians use the HB scale as a quick shorthand alongside a more detailed regional system like the Sunnybrook for tracking changes over time. All three grading systems have shown statistically significant responsiveness to changes when reassessed after three months, indicating they can detect recovery as it happens.⁶

Electrodiagnostic Testing

When the clinical exam tells you the nerve is not working, the next question is whether it is still intact or actively degenerating. This is where electrical testing comes in. The two mainstays are electroneurography (ENoG) and electromyography (EMG), and they answer different questions at different points in the timeline.

ENoG works by delivering a small electrical stimulus to the nerve near the ear and recording the resulting muscle twitch on the affected side, then comparing it to the healthy side. The result is expressed as the percentage of nerve fibers that have degenerated. Clinical guidelines identify ENoG as most valuable between 72 hours and 21 days after the onset of paralysis.⁷ Before 72 hours the nerve fibers have not yet broken down enough to register, and after three weeks the window for certain treatment decisions has passed.

The critical threshold for clinical decision-making is typically set at 90% loss of amplitude. If ENoG shows more than 90% degeneration, the nerve is in serious trouble and surgical decompression may be considered. A loss of 90% or less is classified as a better prognostic sign.⁸ This threshold has become one of the most influential numbers in facial nerve management, though its interpretation depends heavily on the clinical context and timing.

EMG picks up where ENoG leaves off. Rather than stimulating the nerve from outside, EMG places a thin needle electrode directly into the facial muscles to listen for electrical activity. In the early phase of nerve injury, EMG can detect whether any voluntary muscle activity remains, which is especially important in trauma cases because it proves the nerve has not been completely cut. After two to three weeks, EMG becomes the more informative test, because denervated muscles start producing characteristic spontaneous electrical signals called fibrillation potentials, tiny discharges from individual muscle fibers that have lost their nerve supply. Later in recovery, EMG can pick up reinnervation potentials, signaling that the nerve is starting to regrow.⁷

Older techniques like the nerve excitability test (NET) and the maximal stimulation test (MST) are simpler and less expensive than ENoG but also less precise. The NET compares the electrical threshold needed to trigger a muscle twitch on the affected side versus the normal side, while the MST compares the size of the twitch at maximum stimulation. A comparison of the two found that the NET had a prognostic error rate of about 16%, while the MST showed 97% reliability in predicting outcomes.⁹ These tests still see use in settings where ENoG equipment is unavailable, though modern practice guidelines generally favor ENoG combined with EMG.

The Blink Reflex

The blink reflex test bridges the gap between reflex testing and electrodiagnostics. A small electrical tap is applied to the forehead nerve above the eye, and electrodes on the eyelid muscles record the resulting blink. The test produces two responses: R1, an early ipsilateral response traveling a short reflex arc through the brainstem, and R2, a later bilateral response involving a longer pathway. The presence or absence of these components helps distinguish between different lesion sites along the nerve.

The blink reflex also has prognostic value. A retrospective study of facial palsy patients found that the R1 response was a significant independent predictor of recovery, with patients who had a preserved R1 showing more than twice the rate of recovery to near-normal function.¹⁰ The indirect R2 response also contributed independently to predicting full recovery. These findings make the blink reflex a useful early supplement to ENoG and EMG when you want to gauge long-term prognosis.

Why the Cause of Paralysis Changes How You Read the Tests

The same ENoG number can mean very different things depending on what caused the paralysis. Bell’s palsy, the most common form of sudden facial weakness, and Ramsay Hunt syndrome, caused by herpes zoster reactivation, are the two conditions most often evaluated with electrodiagnostics, and they behave differently.

A prospective study comparing the two found that ENoG values were significantly worse in Ramsay Hunt syndrome than in Bell’s palsy at the same stage. The chance of early recovery within six weeks was also significantly lower in Ramsay Hunt syndrome even after correcting for the ENoG effect. Logistic regression showed that the 90% chance of recovery within six months corresponded to roughly 69% degeneration in Bell’s palsy but only about 59% degeneration in Ramsay Hunt syndrome, meaning the latter condition needed a lower threshold of damage to reach the same recovery probability.¹¹ The critical cutoff for non-recovery at one year was about 82.5% degeneration for Bell’s palsy and 78% for Ramsay Hunt syndrome.

Not all research agrees on the precision of these cutoffs. Another study found that recovery rates to near-normal function were 95% in Bell’s palsy and 84% in Ramsay Hunt syndrome, but the authors could not identify a single ENoG value with enough sensitivity and specificity to accurately predict prognosis in either condition.¹² The takeaway is that ENoG is useful as one piece of the puzzle, but no single test result should drive treatment decisions in isolation. Transcranial magnetic stimulation has also been used alongside ENoG in research settings, and one study found that the gap between TMS and ENoG results was smaller in Ramsay Hunt syndrome, suggesting the two conditions affect the nerve at somewhat different locations or through different mechanisms.¹³

Testing for Synkinesis and Aberrant Regeneration

After an episode of facial paralysis, the nerve sometimes regrows imperfectly, sending fibers to the wrong muscles. The result is synkinesis: when you smile, your eye closes involuntarily, or when you blink, the corner of your mouth twitches. The most common complaint is involuntary eye narrowing during mouth movements. A rarer consequence of misdirected regeneration is gustatory hyperlacrimation, sometimes called “crocodile tears,” where the eye tears up whenever the person eats. A modified Schirmer test performed while eating can confirm this specific problem.¹⁴

EMG is particularly helpful for documenting synkinesis. A needle electrode is placed in one facial muscle while the patient deliberately moves a different muscle. If the electrode picks up firing in a muscle that should be resting, synkinesis is confirmed. Multiple-channel recordings, where electrodes sit in several muscles simultaneously, allow clinicians to map which muscle pairs are misfiring and how severely. Surface EMG can also detect it, though needle EMG is more precise for quantification.¹⁴

Synkinesis after facial paralysis needs to be distinguished from hemifacial spasm, a different condition that can look similar but has a different cause and treatment. One way to tell them apart is through neurophysiological monitoring. In hemifacial spasm, a characteristic electrophysiological response called the abnormal muscle response (AMR) can be triggered in all patients, whereas only about 80% of post-paralysis synkinesis patients show an AMR. Additionally, hemifacial spasm produces AMRs with shorter latency and higher amplitude, and the synkinesis pattern differs: in hemifacial spasm, synkinesis occurs sporadically and accompanies involuntary spasms, while in post-paralysis synkinesis it happens with every voluntary movement.¹⁵

Intraoperative Monitoring

When the facial nerve is at risk during surgery, particularly skull base operations or tumor removals near the nerve’s path, real-time monitoring is used to avoid accidental damage. A systematic review of intraoperative neuromonitoring found that free-running and stimulated facial nerve electromyography was the most commonly reported technique, appearing in over 80% of studies. Facial corticobulbar motor evoked potentials were used in about a quarter of studies, and abnormal muscle response monitoring in roughly a fifth. Less common methods like the blink reflex and the Z-L response were also reported.¹⁶

Free-running EMG picks up spontaneous electrical discharges from the facial muscles during surgery. A burst of activity alerts the surgical team that the nerve is being irritated, allowing them to adjust before permanent damage occurs. Stimulated EMG goes a step further by letting the surgeon apply a small probe directly to tissue and see if it triggers a response, essentially asking “is this the nerve?” before cutting. These monitoring tools have become standard of care for procedures like acoustic neuroma removal, parotid gland surgery, and mastoidectomy.

Computer Vision and Automated Assessment

A growing area of research aims to replace or supplement subjective grading with automated systems. The idea is to use computer vision software that analyzes photographs or video of a patient’s facial movements and produces an objective score. One challenge has been that standard facial landmark detection software, trained mostly on healthy faces, performs poorly on faces with paralysis because the expected symmetry is missing. Retraining the software with fewer than 1,600 annotated images of patients with facial palsy significantly improved landmark detection accuracy compared to models trained only on healthy faces.¹⁷

More recent work has used machine learning models that analyze specific facial action units on each side of the face and compare them. A split-face approach that measured mirrored muscle-action intensities achieved accuracy of 83% for the upper face, 93% for the mid-face, and 84% for the lower face, and the severity scores these models produced aligned with those assigned by expert clinicians.¹⁸ These tools are not yet standard clinical practice, but they hold promise for remote monitoring and for creating a more objective record of recovery over time, particularly in settings where a facial nerve specialist is not available.

Patient-Reported Outcomes

Observer-based grading scales measure what a clinician sees, but they miss much of what matters to the patient. Difficulty eating, drinking through a straw, speaking clearly, or feeling comfortable in social situations does not always correlate neatly with a grade on the HB scale. Two validated instruments have been developed to capture the patient’s own experience: the Facial Disability Index (FDI) and the Facial Clinimetric Evaluation (FaCE) Scale. Both are short questionnaires designed to complement observer-based ratings.¹⁹

The FDI divides its questions into a physical function subscale and a social/well-being subscale. Validation work has confirmed that the physical function subscale has good internal consistency and that test-retest reliability for both subscales is solid, with intraclass correlation coefficients above 0.78.²⁰ The social/well-being subscale tends to be a bit less internally consistent, which makes sense given how variable the emotional impact of facial paralysis can be from person to person. Using these tools alongside grading scales gives a fuller picture of how a patient is actually doing, especially when treatment decisions hinge on whether further intervention is worth the risk.

Pediatric Facial Nerve Testing

Evaluating facial nerve function in infants and young children presents unique challenges. Babies cannot follow commands to smile or raise their eyebrows, so assessment relies on observing spontaneous facial movements during crying, feeding, and social interaction. A baby who cries with a perfectly symmetrical mouth but cannot close one eye suggests facial nerve involvement on that side.

Congenital conditions like Moebius syndrome, which involves underdevelopment of the facial nerve and sometimes other cranial nerves, are often diagnosed clinically after other causes of poor feeding and facial immobility have been excluded.²¹ Electrodiagnostic studies can help differentiate Moebius syndrome from other causes of congenital facial weakness. Studies using nerve conduction testing in these children have found that Moebius syndrome and hereditary congenital facial paresis both show low-amplitude facial nerve responses, without the additional nerve or muscle abnormalities that would point to a broader neuromuscular disorder.²² This distinction matters because the treatment path and prognosis differ substantially depending on the underlying diagnosis.