Body temperature can be taken from at least six different sites on (or inside) the body, each with its own normal range and level of accuracy. The most common options for home use are oral, rectal, armpit, ear, and forehead readings. Beyond those, specialized methods like swallowable sensor pills and wearable devices are used in sports, military, and critical care settings. Which method is best depends on the person’s age, the situation, and how precise a reading you need.
Oral Temperature
Placing a digital thermometer under the tongue is the most widely used method for adults and older children. It provides accuracy close to a rectal reading without being invasive, which is why it’s the default choice in most households and doctor’s offices. The normal oral range is 96.4°F to 99.1°F (35.8°C to 37.3°C). For the most reliable result, keep your mouth closed around the thermometer for the recommended time and avoid eating or drinking anything hot or cold for about 15 minutes beforehand.
Rectal Temperature
A rectal reading is considered the gold standard for accuracy because it measures temperature closest to the body’s true core. The normal range runs higher than oral: 98.2°F to 100.8°F (36.8°C to 38.2°C). For newborns through 3 months old, a rectal thermometer is the recommended method because even small inaccuracies matter at that age. It’s also commonly used for toddlers who can’t hold an oral thermometer still. The trade-off is obvious: it’s invasive and uncomfortable, which is why most adults and older children prefer other options.
Armpit (Axillary) Temperature
Taking temperature under the armpit is the least invasive contact method and often the easiest for squirmy toddlers or anyone who can’t use an oral thermometer. The downside is that it’s the least accurate of the common methods. The normal axillary range is noticeably lower, just 96.4°F to 97.3°F (34.8°C to 36.3°C), because the skin surface doesn’t reflect core temperature as closely. If you get an armpit reading that seems like a fever, it’s worth confirming with an oral or rectal reading.
Ear (Tympanic) Temperature
Ear thermometers use an infrared sensor to measure heat radiating from the eardrum, which shares blood supply with the brain’s temperature control center. This makes them fast, usually delivering a result in one to two seconds. The normal tympanic range is 97.0°F to 100.2°F (36.1°C to 37.9°C).
A few things can throw off the reading. Earwax buildup, ear infections, and an unusually shaped or small ear canal can all reduce accuracy. For this reason, ear thermometers aren’t recommended for very young infants. If you use one, make sure the tip is seated properly in the ear canal and pull the ear gently back and up (for adults) to straighten the canal.
Forehead (Temporal Artery) Temperature
Forehead thermometers scan infrared heat from the temporal artery, a blood vessel that runs across the forehead just below the skin. Contactless models became popular during the pandemic because they require zero physical contact. The normal temporal range is 95.4°F to 98.6°F (35.2°C to 37.0°C), a bit lower than oral readings.
Convenience comes at a cost to precision. A clinical study of 265 patients found that contactless infrared thermometers tracked closely with reference devices at normal temperatures but performed poorly when detecting fevers above 99.5°F (37.5°C), catching only about 16% of elevated readings. Direct sunlight, cold air, a sweaty forehead, or holding the device too far from the skin can all skew results further. These thermometers work well for quick screening, but a suspicious reading is worth verifying with an oral or rectal check.
Ingestible Temperature Pills
Originally developed in collaboration with NASA in the 1960s, the ingestible temperature pill is a small capsule (about the size of a large vitamin) containing a tiny sensor, a battery, and a radio transmitter. Once swallowed, the pill measures gastrointestinal temperature every 10 seconds with accuracy within ±0.1°C and transmits a signal to an external recorder worn on the body.
The pill is mostly used in sports science and military research, where athletes and soldiers need continuous core temperature monitoring during intense physical activity. Because it’s inside the body, it captures the real-time rise in core temperature that external methods miss. The person doesn’t feel it once it’s swallowed, and it passes naturally. You do need to swallow the pill at least six hours before the activity to avoid interference from food or fluids. At roughly $40 per single-use pill, it’s far more expensive than a standard thermometer, but it provides a level of continuous, hands-free monitoring that no surface device can match.
Wearable Skin Sensors
Smartwatches and fitness bands increasingly include skin temperature sensors, and researchers have been testing whether these can reliably estimate core body temperature. A systematic review of 25 prediction algorithms found that the best wearable devices can estimate core temperature within about 0.27°C when worn at a single location like the wrist. The most accurate algorithm in the review achieved a margin of just 0.13°C, though that was tested on only 15 subjects.
Accuracy improves significantly when the device combines skin temperature with other data like heart rate and heat flow from the skin, and when environmental conditions (air temperature, humidity) are factored in. Algorithms that included environmental data averaged an error of just 0.18°C, compared to 0.42°C for those that didn’t. These devices aren’t a replacement for a clinical thermometer when you suspect a fever, but they’re useful for tracking trends over time, such as overnight temperature patterns or heat stress during outdoor work.
Invasive Methods in Critical Care
In hospital intensive care units, especially during procedures like therapeutic cooling after cardiac arrest, doctors need extremely precise and continuous core temperature data. The two most common sites are the esophagus (via a thin probe placed through the nose or mouth) and the urinary bladder (via a catheter with a built-in temperature sensor). Esophageal probes respond faster to changes in core temperature, detecting shifts about an hour sooner than bladder probes during cooling procedures. These methods aren’t relevant for home use, but they illustrate why different situations call for different measurement sites.
Why Readings Vary by Site
Every measurement site produces a slightly different number because each reflects a different distance from the body’s warm core. Rectal and esophageal readings sit closest to true core temperature. Oral readings run a bit lower. Armpit and forehead readings are lower still, because they’re measuring skin that loses heat to the surrounding air. There’s no reliable formula to convert one site’s reading to another, so the most useful habit is to always use the same method when tracking your temperature over time. That way you’re comparing apples to apples, and you’ll notice a meaningful change when it happens.