You subtract a degree for ear temperature when you are converting an ear reading to an oral equivalent. Ear (tympanic) thermometers run about 0.5–1.0 °F higher than oral thermometers in most comparisons, so clinicians subtract roughly 1 °F to estimate an oral temperature. You do not add a degree. Adding is a myth inherited from axillary (armpit) readings, which run low and require a small upward adjustment.
Ear thermometers measure infrared energy from the tympanic membrane, which shares blood supply with the brain's temperature-regulating center. That makes tympanic readings fast and reasonably accurate when the probe is placed correctly. The confusion about adding or subtracting comes from mixing up three different sites: oral, axillary, and rectal. Each has its own offset relative to core temperature, and each offset runs in a different direction.
Why the tympanic membrane runs warmer than the mouth
The hypothalamus sits millimeters from the tympanic membrane and receives blood from the same arterial branches. Infrared thermometers aimed at the ear canal detect the thermal radiation of that membrane and the surrounding canal wall. Because the membrane reflects core temperature closely, a well-placed ear reading lands near core, and core typically sits slightly above oral temperature.
Oral thermometers, by contrast, measure a sublingual pocket that cools with mouth breathing, drinking, and ambient air. That pocket runs roughly 0.5–1.0 °F below core in steady conditions. The gap between ear and oral is therefore mostly an artifact of the oral site running cool, not the ear site running hot. When you subtract a degree from an ear reading to match an oral reading, you are correcting for the mouth's cool bias.
The physics matter here. Infrared sensors do not touch tissue; they measure emitted radiation proportional to the fourth power of absolute temperature. A probe that misses the membrane and reads canal wall will underestimate core. A probe pressed against wax will do the same. Placement quality, not the conversion rule, drives most real-world error.
Quantified offsets across measurement sites
Published comparisons place tympanic readings near core, oral readings slightly below, axillary readings clearly below, and rectal readings at or slightly above core. The offsets are approximate and vary by device, age, and technique. Treat them as typical ranges, not fixed constants.
| Site | Typical offset vs core | Direction to convert to oral | Notes |
|---|---|---|---|
| Tympanic (ear) | Near core, about 0.5–1.0 °F above oral | Subtract about 1 °F | Accuracy depends on probe placement and wax |
| Oral (sublingual) | About 0.5–1.0 °F below core | Reference site | Affected by drinking, mouth breathing |
| Axillary (armpit) | About 1–2 °F below core | Add about 1 °F to compare with oral | Least accurate; common source of the "add a degree" rule |
| Rectal | At or slightly above core | Subtract to match oral | Reference standard in young infants |
| Temporal artery (forehead) | Variable, often near oral | Device-specific | Follow manufacturer calibration |
The "add a degree" habit traces to axillary thermometry, where the armpit runs cool and a small upward correction approximates oral. People carried that rule to ear readings, where it points the wrong way. Ear readings already sit at or above oral, so adding inflates them further. The correct direction for ear-to-oral conversion is subtraction.
One counterintuitive point: the offset is not a fixed property of the ear. It is a property of the comparison. If your reference is core temperature, the ear reading needs little or no correction. If your reference is oral, subtract. If your reference is axillary, the gap is larger. The "degree" people argue about is really a site-to-site bridge, not an inherent ear correction.
Where the one-degree rule breaks down
The subtraction rule is a rule of thumb, and thresholds shift with age, device, and technique. In newborns and young infants, rectal thermometry has historically been the reference standard, and ear readings are less reliable because the ear canal is small and curved. Many pediatric guidance bodies advise caution with ear thermometers under about 6 months of age. For that group, an ear reading should be confirmed with a rectal or clinical assessment rather than adjusted by a fixed degree.
Device calibration adds another variable. Different infrared thermometers use different algorithms and emissivity assumptions. Two devices on the same person can disagree by a degree or more. Manufacturer instructions sometimes specify a site-specific mode or offset; follow those instructions over generic rules. A single subtraction constant cannot correct for a poorly calibrated device.
Ambient conditions also matter. A cold ear canal after outdoor exposure, earwax impaction, or recent ear drops can shift readings. Otitis media (ear infection) can raise local readings. These effects push individual readings around the nominal offset, which is why a single degree is an approximation and not a clinical conversion factor.
Red-flag warnings: Any fever in an infant under 3 months warrants prompt medical evaluation, regardless of the site used. A reading that seems inconsistent with how the person looks should be rechecked at another site. Suspected ear infection with fever should be assessed clinically, not managed by adjusting numbers. Very high readings (for example, above 104 °F / 40 °C) call for urgent care, not arithmetic. Repeated readings that swing widely suggest technique or device problems, not true temperature change.
Common misconceptions about ear temperature
The biggest misconception is that ear thermometers are inherently inaccurate. Well-placed tympanic readings track core reasonably closely; the failure mode is technique, not the site. A probe that does not seal the canal or aims at wax produces a low reading, which people then "correct" by adding a degree, compounding the error.
A second misconception is that the "add a degree" rule is universal. It is not. It belongs to axillary thermometry. Applying it to ear readings inflates estimates and can make a normal temperature look like a low-grade fever.
A third misconception is that ear and oral readings should be identical. They measure different things. Expecting them to match exactly ignores the physiology of each site. The subtraction adjusts the ear reading toward an oral equivalent for comparison, not because one is wrong.
A fourth is that fever thresholds are the same everywhere. Clinical fever definitions are usually anchored to a specific site, often rectal or oral. If you convert an ear reading, you are approximating that anchor, and the approximation carries uncertainty. For screening, trends and clinical appearance matter as much as the number.
Practical steps for taking and interpreting ear temperatures
Use these steps to get readings worth converting. Confirm the device is set to ear mode and has fresh batteries or a clean probe cover. Straighten the ear canal: gently pull the ear back and up for adults and older children; back and down for young children. Seat the probe snugly so it seals the canal; a loose seal reads low. Wait 15–30 minutes after outdoor cold exposure or a hot drink before measuring. Take two or three readings and use the highest consistent value. Record the site and device alongside the number so trends stay comparable.
For interpretation, use the subtraction only when you need an oral-equivalent number for a threshold that is defined orally. If your clinician gave a threshold in ear terms, do not convert. Keep one reference site per person over an illness course, because mixing sites creates false trends. When in doubt, recheck at the same site with correct technique before adjusting anything.
A practical threshold rule: if an ear reading is at or above the fever cutoff you were given for oral measurement, subtract about 1 °F and compare. If the adjusted value still crosses the threshold, treat it as a fever signal and confirm clinically. If the adjusted value lands just below, recheck rather than assume normal.
Limits of the one-degree correction
The subtraction is a population-level approximation, not a personal calibration. Individual offsets vary with anatomy, age, wax, and device. No home user can derive a precise personal offset without a reference measurement, and clinical thermometry guidance generally treats ear readings as screening tools rather than definitive core measurements.
Devices and standards also change. Regulatory clearance and manufacturer algorithms differ, and guidance on preferred sites in infants has shifted over time. Where current clinical recommendations matter, defer to your clinician and the device's instructions rather than a generic conversion constant.
Finally, numbers do not replace judgment. Temperature is one signal among many. Clinical appearance, symptom duration, and age drive decisions more than a single adjusted reading. Use the subtraction to make comparisons fair, then act on the whole picture.