Where to Use an Infrared Thermometer on the Body

The forehead is the most widely recommended site for a non-contact infrared thermometer, and for most everyday purposes it gives a reasonable estimate of whether someone has a fever. But “reasonable” is doing a lot of work in that sentence. Infrared thermometers read skin surface temperature, not true core body temperature, and the gap between those two numbers shifts depending on where you point the device, how recently the person was outdoors, and even what they have on their skin. The forehead became the default during mass fever screening, yet research on alternative sites like the ear canal, neck, and wrist tells a more complicated story about accuracy, convenience, and when each site actually makes sense.

The Forehead as the Standard Measurement Site

Most non-contact infrared thermometers (NCITs) sold for home or screening use are designed and calibrated for the center of the forehead. The forehead’s appeal is practical: it is easy to access, requires no physical contact, and the superficial temporal artery runs just beneath the skin there, making skin temperature a rough proxy for core temperature. In a study comparing infrared devices against a validated core-body-temperature sensor, an infrared gun held about 3 cm from the forehead showed the narrowest range of error among several body sites, with readings averaging only about 0.2°C above core temperature and most individual readings falling within roughly a degree of the reference value.¹ A separate study of postoperative patients found that forehead infrared readings correlated well with core temperature and were comparable to tympanic (ear) readings in accuracy.²

A meta-analysis pooling data from over 5,500 readings across ten different NCIT devices found that when measurement was restricted to the forehead, pooled sensitivity for detecting fever was about 81% and specificity was about 92%.³ That means roughly one in five people with a genuine fever will read as normal. For home use when you just want to know if your child feels warm because they are actually running a fever, that is usually good enough. For clinical situations where missing a fever has serious consequences, it falls short.

The Ear Canal and Why It Sounds Better Than It Performs

Infrared tympanic thermometers, which you insert into the ear canal to read heat radiated from the eardrum, were once considered a major step forward. The tympanic membrane shares blood supply with the brain’s temperature-regulation center, so in theory it should give a reading close to core temperature. In practice, results are mixed. A systematic review found that tympanic readings in children averaged about 0.2°C below rectal temperature, which sounds small, but the range of individual disagreement was wide enough that any single reading could be off by more than a degree in either direction.⁴ A Lancet systematic review reached a similar conclusion: average differences between ear and rectal readings were small, but the wide limits of agreement meant ear temperature was not a reliable substitute when precision mattered.⁵

Fever detection is where tympanic thermometers really struggle. One systematic review estimated that infrared ear thermometers would miss fever in three or four out of every ten febrile children when rectal temperature was used as the benchmark.⁶ That is a troubling miss rate for a device marketed as convenient and reliable.

Part of the problem is anatomy. The shape and size of the ear canal affects how much of the tympanic membrane the sensor can actually “see.” Research has shown that wider ear canals and a clear line of sight to the eardrum produce more accurate readings, while narrow or curved canals introduce more error. About 22% of the variation in reading accuracy was explained by ear canal shape alone.⁷ Earwax makes things worse. A pilot study found that cerumen blocking the canal tended to lower readings, with nearly half of participants showing a drop of at least 0.3°C.⁸ If you do use an ear thermometer, a clean, dry ear canal pointed straight at the eardrum gives the best chance of an accurate result.

The Neck, Wrist, and Other Alternative Sites

During mass screening events, especially at building entrances during the COVID-19 pandemic, screeners sometimes pointed infrared thermometers at the wrist, neck, or temporal region instead of the forehead. These sites vary widely in how well they track core temperature.

The neck consistently outperforms other alternative sites. A study comparing forehead, temporal, wrist, and neck readings against axillary temperature found that neck temperature was the most consistent with the reference, with the highest area under the curve for fever detection. At an optimal cutoff of about 37.35°C, neck readings achieved roughly 87% sensitivity and 85% specificity.⁹ Forehead, temporal, and wrist temperatures all required significantly lower cutoff thresholds to flag fever, which is a sign they run cooler than core temperature and need recalibration to be useful at those sites.

The wrist became popular in some Asian screening protocols because it is easily exposed and less affected by ambient temperature swings. A prospective study found that wrist readings were more stable across different environments than forehead readings, though both sites showed meaningful gaps compared to tympanic temperature. Wrist measurements ran about 0.6 to 1.0°C below the ear-based reference, while forehead readings ran 0.6 to 1.7°C below.¹⁰ Wrist screening can flag fever at a population level, but the temperature you see on the display will look deceptively low compared to what a traditional thermometer would show. Any wrist-based screening protocol needs a lower fever threshold, around 36.2°C, to catch what a rectal or oral thermometer would call 37.3°C or higher.

A study conducted in an African outpatient clinic paints a grimmer picture for contactless thermometers used in real-world tropical conditions. The non-contact thermometer detected only 13% of patients with a fever above 38°C.¹¹ That is barely better than flipping a coin, and it highlights how dramatically performance can vary depending on device quality, calibration, and environment.

Children Are Harder to Read

Parents tend to reach for an infrared thermometer because it is fast, non-invasive, and does not wake a sleeping child. These are real advantages, but accuracy in children deserves a closer look. A systematic review and meta-analysis of NCITs found that sensitivity for detecting fever was higher in children than in adults (about 79% versus 70% overall), which is somewhat reassuring.¹² A separate meta-analysis focused on infrared tympanic thermometers in children older than one year found pooled sensitivity of about 80% and specificity around 94%.¹³

Those numbers are decent for a screening pass, but they still mean roughly one in five febrile children will get a normal reading. For a child who looks unwell, a single normal infrared reading should not be the final word. Pediatricians often recommend following up with a rectal thermometer in infants and very young children when the clinical picture does not match the number on the screen. Infrared thermometers are best understood as a first-pass tool: convenient for quick checks, but not a definitive replacement for contact thermometry when the stakes are high.

What Throws Off the Reading

Infrared thermometers measure the thermal radiation your skin emits, then use an internal algorithm to convert that into an estimated body temperature. Anything that changes your skin surface temperature without changing your core temperature will introduce error. Several of these factors are common enough that they trip people up regularly.

Cold exposure is the most obvious culprit. If someone has just come in from a cold day, their forehead may read well below their actual core temperature. A pediatric study found that children who had been exposed to cold outdoor conditions produced contradictory results until they had acclimated indoors for at least ten minutes.¹⁴ The practical lesson: wait at least ten minutes after someone comes inside before scanning their forehead.

Exercise and sweating create the opposite problem. During physical activity, skin temperature can temporarily overshoot core temperature as blood rushes to the surface for cooling. Research found that a forehead infrared thermometer significantly overestimated core temperature about fifteen minutes after the start of exercise, and did not give reliable readings during any period of rising body temperature.¹⁵ The device worked fine under stable resting conditions, but fell apart when the body was actively heating up or cooling down. Sweat compounds this problem. Research on exercising athletes confirmed that wind and skin sweat together invalidate infrared skin thermometry during exertion, regardless of which device brand is used.¹⁶ The evaporative cooling from sweat drops the apparent skin temperature, while the underlying skin remains warm. The sensor reads the sweat surface, not the skin beneath it.

Cosmetics, Sunscreen, and Other Surface Barriers

This is one that most people never think about. Anything applied to the forehead can change the thermal reading an infrared sensor picks up. A study on the thermal effects of common topical products found that hydroalcoholic gel (the kind used as hand sanitizer) dropped forehead skin temperature by more than 2°C within one minute of application. Organic sunscreens produced a drop of up to 1.7°C. Recovery took about nine minutes.¹⁷ That is easily enough to mask a real fever.

Color cosmetics with mineral pigments like iron oxide, and sunscreens with mineral filters like zinc oxide or titanium dioxide, had little to no effect on thermal readings. The key variable is whether the product has an evaporative component. Alcohol-based gels and sprays cool the skin surface through evaporation, creating a false reading. Cream-based or powder-based products that sit as a stable layer tend to be thermally neutral. If you have just applied hand sanitizer and it splashed or dripped onto your forehead, or if you are wearing an alcohol-based setting spray, wait a few minutes before taking a reading.

Does Skin Tone Affect Accuracy

This question gained urgency during the pandemic, when thermal screening was deployed at scale across diverse populations. The physics involved are worth understanding at a basic level. Infrared thermometers rely on the skin’s emissivity, which is a measure of how efficiently a surface emits thermal radiation. Human skin of all tones has high emissivity in the long-wave infrared range that medical thermometers use, generally around 0.98. Melanin absorbs and emits visible and near-infrared light differently depending on concentration, but in the wavelength range medical devices operate in, the effect should theoretically be negligible.

The real-world data is more complicated. A study using thermal imaging found a statistically significant interaction between melanin level and measured temperature change. The measured difference between skin tone categories was about 0.33°C per category, adding up to roughly a full degree Celsius of discrepancy between lighter and darker intermediate skin tones under the same cooling conditions.¹⁸ However, a separate study measuring erythema-related temperature across skin tones found no significant difference in temperature change between groups, though it cautioned that high variability in results made it premature to conclude thermal imaging works equally well across all skin tones.¹⁹

The honest assessment is that we do not yet have definitive evidence that skin tone introduces clinically dangerous levels of error in standard fever screening, but we also cannot rule it out. The existing studies point in slightly different directions, and most have small sample sizes. If you have darker skin and you are consistently getting readings that feel low relative to how you feel, it is worth cross-checking with a contact thermometer rather than assuming the infrared device is right.

Not All Devices Are Created Equal

One of the most underappreciated problems in infrared thermometry is device quality. The pandemic created enormous demand for non-contact thermometers, and manufacturers of widely varying quality rushed to fill it. A clinical evaluation of multiple NCIT models found that depending on the model, 48% to 88% of individual readings fell outside the manufacturer’s own stated accuracy range. The sensitivity for detecting a temperature above 38°C ranged from zero with one model to 0.69 with the best-performing model.²⁰ A device with zero sensitivity for fever detection is, functionally, a random number generator with a digital display.

Some of this variability comes down to calibration. Many consumer-grade NCITs are advertised with accuracy claims of plus or minus 1°C or 2°C, which is far too loose for meaningful fever detection. Devices specifically calibrated for human skin temperature ranges between 30°C and 40°C can achieve accuracy within 0.3°C, but these tend to be medical-grade instruments with a price tag to match. If you are buying an infrared thermometer for home use, look for one that meets the ASTM E1965 or IEC 80601-2-59 standards, which specify tighter accuracy requirements for clinical-grade skin thermometers. A cheap unbranded device from an online marketplace may technically be an infrared thermometer, but that alone does not mean it will give you a meaningful reading.

Practical Tips for Getting the Best Reading

Given all these variables, a few straightforward habits can improve your results:

• Wait after coming indoors: At least ten minutes of indoor acclimation before scanning, especially in cold weather.
• Dry, product-free skin: Wipe away sweat, and avoid scanning over areas with recently applied sanitizer, sunscreen, or alcohol-based cosmetics.
• Rest before measuring: Take readings at rest, not right after exercise or a hot shower. The device cannot track a moving target.
• Correct distance: Hold the thermometer at the distance specified in its manual, typically 3 to 5 cm from the center of the forehead. Too close or too far changes the reading.
• Take multiple readings: A single scan is a snapshot with noise. Two or three readings averaged together give a more reliable number.
• Know your baseline: Scan yourself a few times when you feel healthy at different times of day. Body temperature follows a daily rhythm, running lower in the morning and higher in the late afternoon. Knowing your personal normal range helps you interpret a borderline reading.

When an Infrared Thermometer Is Not Enough

Infrared thermometers are screening tools, not diagnostic instruments. They are good at telling you quickly whether someone might have a fever, especially in settings where speed and no-contact measurement matter. They are not good at telling you a person’s precise core temperature. For situations where accuracy truly matters, such as monitoring a newborn, tracking a fever in a hospitalized patient, or evaluating someone with symptoms of serious infection, contact thermometry remains the standard. Rectal temperature is still considered the clinical gold standard for core body measurement, particularly in young children. Oral and axillary readings from contact thermometers, while not perfect, consistently outperform non-contact infrared devices for precision.

A study in a low-resource tropical setting illustrated this gap starkly. Axillary thermometry, despite its own known limitations, detected fever more than twice as reliably as non-contact infrared devices when the fever threshold was adjusted appropriately.²¹ The infrared devices were faster and more accepted by patients, but their clinical accuracy lagged behind. That tradeoff between convenience and precision is the central tension of infrared thermometry, and understanding it is ultimately more useful than any specific recommendation about where on the body to point the device.