What Is Considered a Fever With an Ear Thermometer?

A temperature of 38.0 °C (100.4 °F) or higher on an ear (tympanic) thermometer is widely considered a fever in both adults and children. That number aligns with the rectal threshold because the ear canal sits close to the blood supply that feeds the brain, giving it a reading near true core body temperature. The story gets more interesting, though, when you look at what “normal” ear temperature actually is across different ages, what can nudge the reading up or down, and when this tool is trustworthy versus when you should reach for a different method.

Why the Ear Threshold Matches Rectal, Not Oral

If you have ever compared fever charts for different thermometer types, you have probably noticed that oral thermometers use a lower cutoff, typically around 37.5 °C (99.5 °F), while rectal and ear thermometers share the 38.0 °C line. The reason is anatomy. The tympanic membrane sits just millimeters from the internal carotid artery, one of the main vessels carrying warm blood to the brain. An infrared ear thermometer reads the heat radiating from that membrane and the ear canal, so in theory it is capturing something close to your core temperature rather than the slightly cooled surface temperature you get under the tongue.

A cross-sectional study of hospitalized adults found that ear readings ran about 0.1 to 0.2 °C lower than rectal readings on average, and the researchers concluded that at a screening cutoff of 37.5 °C the ear thermometer was accurate enough to catch fever reliably, while higher cutoffs risked missing genuinely febrile patients.¹ That small gap means using the standard 38.0 °C cutoff with an ear thermometer is reasonable, but a reading of 37.8 °C in someone who looks unwell still deserves attention. The number is a guideline, not a cliff edge.

How Close Are Ear and Rectal Readings, Really?

Rectal temperature remains the reference standard in clinical research, so every other method gets judged against it. In a study of infants and toddlers, the average rectal temperature was 37.58 °C and the average ear temperature was 37.60 °C, almost identical overall. But the agreement was not perfect across the range: when rectal readings were on the low end, the ear tended to read a bit high, and when rectal readings were elevated, the ear tended to read a bit low.² In practical terms, that pattern means ear thermometers are slightly more likely to underestimate a high fever than to miss a low-grade one.

A separate accuracy study comparing multiple thermometry devices against an ingested capsule sensor (the current gold standard for core temperature) found that a well-regarded tympanic thermometer had a mean error of just 0.044 °C, the smallest of any device tested. A sign test showed it was the only device whose readings were not statistically different from the gold standard.³ That is impressive, but the 95 percent limits of agreement still ranged from about −0.8 °C to +0.7 °C, which means any individual measurement could be nearly a full degree off. For screening purposes that is fine; for decisions like whether to rush to the emergency department, it is worth measuring more than once.

What “Normal” Looks Like at Different Ages

One reason a single fever cutoff can mislead you is that baseline ear temperature is not the same for everyone. A study that measured ear temperature across a broad age range found that the overall mean was about 36.4 °C, but the number shifted with age: adolescents averaged 36.5 °C, adults averaged 36.4 °C, and people over 65 averaged only 36.1 °C.⁴ That half-degree gap between a teenager and a grandparent matters. An older adult with an ear reading of 37.8 °C may effectively be running a fever even though the number falls just under the standard 38.0 °C line, because their personal baseline is so much lower.

The same study also found a small sex difference: women averaged about 36.5 °C while men averaged 36.3 °C. These are population averages, and individual variation is considerable, but the takeaway is that knowing your own (or your child’s) normal baseline temperature gives you more useful information than relying solely on a universal cutoff.

Young Infants

Ear thermometers are popular with parents of small children because they are fast and relatively non-invasive. For children roughly six months and older, the evidence supports their use as a screening tool. In a large pediatric study of nearly a thousand children, the ear thermometer detected fever with high accuracy, and it was the only non-invasive device that reached a sensitivity above 90 percent in children aged six months to five years.⁵ Below six months, the picture changes. Infants’ ear canals are extremely narrow and curved differently, making it harder to aim the sensor at the tympanic membrane. Research on infants three months and younger found that ear thermometers performed well when the baby was not febrile but were less reliable at catching actual fevers in that age group.⁴ Most pediatric guidelines still recommend rectal thermometry for babies under three to six months when fever detection is critical.

Older Adults

At the other end of the age spectrum, the same pediatric study found that fever detection sensitivity dropped in the oldest children tested. In geriatric patients more broadly, the lower baseline noted above means that a “normal” ear reading can coexist with a real infection. Additionally, older adults are more likely to have significant earwax buildup, which introduces another source of error discussed below. If you are monitoring temperature in someone over 65, consider treating readings in the 37.5 to 37.9 °C range with more suspicion than you would in a younger adult.

Factors That Can Skew an Ear Reading

Even a well-designed tympanic thermometer is only as good as the conditions under which you use it. Several everyday factors can push the reading up or down by enough to matter.

Earwax

A buildup of cerumen (earwax) partially blocks the infrared sensor’s line of sight to the tympanic membrane, and the wax itself is cooler than the membrane. A study of older adults found that the presence of cerumen lowered the reading by an average of about 0.13 °C, with some individuals seeing a drop of 0.5 °C or more before and after wax removal.⁶ Nearly half the subjects in that study had readings lowered by at least 0.3 °C. If you know you or the person you are measuring has heavy earwax, the thermometer may underestimate a fever. Consider switching to a different site or having the wax removed before relying on ear readings for clinical decisions.

Lying on the Ear

If you have been sleeping on one side, the ear that was pressed against the pillow will be warmer for a while. Research has confirmed that tympanic readings are significantly higher when the measurement is taken from an ear that was recently lain on compared to one that was not.⁷ The practical fix is simple: use the ear that was facing up, or wait a few minutes after the person gets up. In a hospital setting, nurses are generally trained to check which side a patient was lying on, but at home this detail is easy to forget, especially when checking a sick child in the middle of the night.

Technique and Positioning

The sensor needs a clear path to the eardrum. In adults, gently pulling the outer ear up and back straightens the ear canal; in young children, you pull it down and back. If the probe is aimed at the canal wall rather than the membrane, the reading can be falsely low. Most modern ear thermometers beep or give an error if they do not detect a proper seal, but cheaper models may simply return a number without warning you that the angle was off. Taking two or three readings in the same ear and using the highest one is a reasonable hedge against a single bad measurement.

Does an Ear Infection Affect the Reading?

Parents often wonder whether measuring temperature in an infected ear gives a misleadingly high number. The short answer is that it raises the reading slightly, but probably not enough to change any clinical decision. A study of children with one-sided (unilateral) ear infections found that in suppurative otitis media, where pus was present behind the eardrum, the infected ear read about 0.38 °F (roughly 0.2 °C) higher than the uninfected ear. When neither ear had pus, the difference was not statistically significant. And even the infected ear’s reading did not differ meaningfully from oral or rectal control measurements.⁸

A more recent study found a slightly larger gap, about 0.5 °C, between the infected and uninfected ear in children with acute otitis media, and suggested the difference might even be useful as a diagnostic clue for ear infections in children with unexplained fevers.⁹ The practical advice is straightforward: if you suspect an ear infection on one side, measure the other ear. If both ears are involved, the reading may be slightly elevated by the local inflammation, but the bump is small enough that it should not prevent you from detecting or ruling out a systemic fever.

Ear Thermometers vs. Forehead Thermometers

Forehead (temporal artery) thermometers gained popularity during the COVID-19 pandemic because they require no contact. How do they stack up against ear thermometers? In the large pediatric study mentioned earlier, the ear thermometer had a significantly greater ability to detect fever than the temporal thermometer. The ear device’s agreement with rectal temperature was also tighter, with 95 percent of readings falling within about 1 °C of rectal, compared to a wider spread of up to 1.5 °C for the forehead device.⁵ An evidence-based nursing review similarly concluded that a well-known ear thermometer model was more reliable than a widely used temporal scanner for assessing fever in children.¹⁰

That said, one postoperative study in adults found that both forehead and tympanic thermometers correlated well with core temperature and were considered comparable for use in the early recovery period.¹¹ The gap between the two devices seems to matter most in children and in situations where the ambient temperature is variable. Forehead readings are more susceptible to environmental conditions, such as outdoor cold or sweating, than ear readings, which are sheltered inside the canal. A study of temperature measurement during outdoor exercise found poor agreement between measurement sites, with standard deviations in the comparisons exceeding 0.67 °C, concluding that common methods should not be treated as interchangeable during physical activity.¹²

Armpit Readings and Why They Disagree

Axillary (armpit) thermometers are cheap and noninvasive, which makes them a common backup. But armpit temperature is a surface measurement, and it consistently runs lower than ear temperature. A study comparing a wearable axillary sensor with a tympanic thermometer in children found a mean difference of about 0.45 °C, with the armpit reading lower. The 95 percent limits of agreement stretched from −0.57 °C to +1.46 °C, meaning individual readings could be off by more than a full degree.¹³ If you typically use an armpit thermometer and get a borderline reading, switching to an ear thermometer for a more reliable number is a sensible step.

Calibration and Device Quality

Not all infrared thermometers are created equal. A study comparing calibrated and non-calibrated handheld infrared thermometers against a calibrated digital thermometer in both healthy and febrile adults found significantly more variance in the non-calibrated devices. The calibrated instruments showed tighter agreement with each other, while the uncalibrated ones introduced additional scatter that could matter at the boundary between “warm” and “fever.” In the febrile group, all three thermometer types correlated more strongly with each other, which makes sense: a high fever produces a bigger signal that even a less precise instrument can detect. The concern is at the margins, right around that 38.0 °C line, where device quality and calibration make the difference between catching a low-grade fever and missing it.

If you are buying an ear thermometer for home use, look for one that meets medical-device standards in your country. Devices marketed purely as consumer electronics sometimes skip the clinical validation that medical-grade models go through. The difference in cost is usually modest, and the difference in reliability at borderline temperatures can be meaningful.

When to Measure and How Often

Body temperature naturally fluctuates throughout the day. It tends to be lowest in the early morning and peaks in the late afternoon or early evening, with a swing of about 0.5 °C in most people. If you are tracking a fever, try to measure at roughly the same times each day so you are comparing like with like. A reading of 37.9 °C at 7 a.m. is more concerning than the same reading at 5 p.m., because morning temperatures run cooler.

For a child who feels warm, a single reading above 38.0 °C on an ear thermometer is enough to call it a fever and start thinking about whether to consult a doctor, give a fever reducer, or simply monitor. You do not need to confirm with a second type of thermometer unless the reading seems inconsistent with how the child looks and acts. A toddler running around the room with an ear reading of 39.2 °C may warrant a re-check; a listless child with 38.3 °C does not need additional proof.

For adults monitoring their own temperature, the same 38.0 °C threshold applies. In older adults with a lower baseline, a reading of 37.5 °C that is accompanied by chills, confusion, or unusual fatigue deserves the same attention as a higher number in a younger person. The thermometer gives you a data point; context turns it into a decision.

Left Ear vs. Right Ear

You might assume both ears give the same reading, and for most practical purposes they do. But small differences between the two sides are common, driven by slight differences in blood flow, canal geometry, and even which side of the head is warmer at any given moment. The accuracy study that found exceptional performance from a tympanic thermometer noted that the left ear readings were more accurate than the right ear readings using the same device, with the left ear achieving the best agreement with the gold standard.³ Whether this was due to anatomy, handedness affecting probe angle, or chance in that particular sample is hard to say. Some clinicians recommend consistently using the same ear each time you measure so that trends over time are not confounded by switching sides. If one ear has significant wax buildup or an infection, of course, use the other one.

The broader lesson from side-to-side variability, earwax effects, positional warming, and device differences is that a single temperature reading is a snapshot with some noise baked in. The useful pattern comes from multiple readings over time, not from treating any one number as sacred. A rising trend from 37.5 to 38.2 to 38.8 °C tells you more than a lone 38.1 °C ever could.