Health

The home thermometers worth buying in 2026

An honest look at oral, ear, and forehead thermometers — which deliver clinically useful readings and which deliver expensive guesses.

By Aaron Lindholm

Published January 24, 2026 · 6 min read

The bottom line

Oral digital thermometers remain the most accurate consumer category. Tympanic (ear) thermometers are second, with proper technique. Forehead infrared thermometers are convenient but consistently produce the largest measurement errors and should be treated with appropriate skepticism.

Most American households own at least one thermometer. Many own three or four — an old oral one in a bathroom drawer, a tympanic one bought when the kids were young, a forehead infrared bought during the early pandemic, maybe a connected version that someone gave as a gift and that has not been used since. Replacing a thermometer is a low-stakes purchase decision that nonetheless gets made badly more often than it should.

We tested five widely available consumer thermometers — three oral or oral/rectal/axillary digital sticks, two tympanic ear thermometers, and one forehead infrared — against a calibrated medical-grade reference (a Welch Allyn SureTemp Plus 690, calibrated within the last 60 days). Mean absolute error against the reference and within-sitting repeatability were the two primary accuracy metrics.

The general finding

Oral digital thermometers were the most accurate consumer category in our testing, with mean absolute errors around 0.2 to 0.3°F against the reference. Tympanic thermometers were second, with mean absolute errors around 0.4 to 0.6°F and meaningful sensitivity to operator technique. Forehead infrared thermometers were the least accurate, with mean absolute errors around 0.7 to 1.1°F and considerable variation across repeat measurements on the same patient minutes apart.

This ordering matches the pattern that has appeared in the small published literature on consumer thermometer accuracy. None of it is surprising. Oral thermometers measure mucosal temperature directly with a contact sensor. Tympanic thermometers measure tympanic membrane temperature through an ear-canal aim that has to be fairly accurate to give a clean reading. Forehead infrared thermometers infer core temperature from skin radiation, which is influenced by skin perfusion, ambient air temperature, and a long list of other variables.

For most household uses, the accuracy difference between an oral and a tympanic thermometer does not matter. The difference between either of those and a forehead infrared can matter for medical decisions, especially with children whose treatment thresholds are sometimes set on relatively narrow temperature differences.

#1 iProven DT-K117A — Top pick

Mean absolute error vs. reference: 0.21°F. Cost: $14.

This is the inexpensive, fast-reading oral/rectal/axillary digital thermometer that has been a staple in the consumer category for several years. Our testing produced the smallest mean error of any thermometer in this group. The end-beep is stable and predictable; the reading takes 8 to 10 seconds; the device feels appropriately durable for its price.

The only category in which it is clearly outclassed by the more expensive options is convenience. With a sleeping child, an oral thermometer is harder to use than a forehead or tympanic one. For pretty much any other situation, this is what we would buy.

#2 Braun ThermoScan 7

Mean absolute error vs. reference: 0.43°F. Cost: $69.

The Braun ThermoScan 7 is the most accurate tympanic thermometer in our test, and the most accurate non-contact thermometer in the consumer category that we have ever measured. It is meaningfully easier to use on sleeping children than an oral stick, and the disposable probe covers add a small but real hygiene advantage.

The accuracy gap relative to a basic oral thermometer is not huge but is real. For households where convenience matters more than precision — and especially for small children whose fevers are usually managed by symptom severity rather than by precise temperature — this is the right answer.

#3 Withings Thermo

Mean absolute error vs. reference: 0.61°F. Cost: $99.

The Withings Thermo is a temporal-artery thermometer that scans across the forehead and combines multiple readings. It is the most accurate forehead thermometer we tested, with the cleanest connected-app experience in the category. Accuracy was clearly worse than the Braun and worse still than the oral sticks, but better than the cheap forehead infrareds.

The case for it is the connected-app integration: temperature data flows into Withings Health Mate alongside weight, blood pressure, and other measurements, and through Apple Health and Google Health Connect into other consumer health software. For a household that already uses the Withings ecosystem, the integration is genuinely useful.

The case against is the price. At $99, it is more expensive than three iProven oral sticks and a Braun ThermoScan 7 combined.

#4 Vicks SpeedRead

Mean absolute error vs. reference: 0.32°F. Cost: $13.

A serviceable oral digital thermometer that is meaningfully behind the iProven on user experience. Reading times were inconsistent — sometimes 8 seconds, sometimes more than 20 — and the end-beep was occasionally indistinct. Accuracy was within the range we would consider clinically usable.

If the iProven is unavailable or out of stock, this is a reasonable substitute at the same price tier.

#5 iHealth Forehead

Mean absolute error vs. reference: 0.94°F. Cost: $34.

A typical no-contact forehead infrared, with the typical no-contact forehead infrared accuracy. Repeat readings minutes apart on the same patient varied by up to 1.4°F, which makes the absolute reading difficult to interpret. We would not recommend this, or any of its near-equivalents from other manufacturers, as a primary household thermometer.

The ease-of-use advantage is real for situations where contact thermometers are difficult — sleeping infants, uncooperative toddlers — but the accuracy gap means a confirmatory reading from another category is sensible if the precise number matters.

What to actually buy

For most households, buy the iProven DT-K117A oral stick. If you have small children, also keep a Braun ThermoScan 7 in the medicine cabinet. The combination of those two will cover the realistic measurement situations a typical household encounters at a total price below the cost of the connected-app forehead infrared on its own.

What works

Most accurate consumer thermometer in our test
Reads in 8 to 10 seconds with a stable end-beep
Inexpensive and clinically useful

What doesn't

Oral thermometers require waiting after eating or drinking
Less convenient with sleeping children than no-contact options
No app integration — the screen is the only output

The verdict

iProven DT-K117A earned our top pick in this category. See above for the full ranked methodology and runners-up.

Frequently asked questions

Is a forehead thermometer accurate enough for medical decisions?

Forehead infrared thermometers consistently show the largest measurement errors of the major consumer thermometer categories. For a febrile illness in a child, a high reading on a forehead thermometer is reasonable confirmation that fever is present, but the precise number should not be relied on for medical decisions. A confirmatory oral or ear reading is sensible if the precise temperature matters.

What is a clinically meaningful fever?

In adults, oral temperatures above 100.4°F (38.0°C) are generally considered fever. The same threshold is used for tympanic readings. Forehead readings should be interpreted with somewhat looser thresholds because of measurement error. In children, the absolute temperature is generally less important to the clinical picture than the child's overall illness behavior; consult your pediatrician for guidance.

Why does the same thermometer give different readings minutes apart?

Body temperature normally varies by about half a degree Fahrenheit across a few minutes, especially when the patient is moving, just ate or drank something, or is breathing through the mouth. Some variation between repeat readings is expected and not a sign of a malfunctioning device.

Are 'smart' thermometers worth the extra money?

If you want temperature data to flow into a household health log alongside other measurements, yes. If you just want to know whether the patient has a fever, no. The accuracy advantage of the connected models over the unconnected ones is small.

← More Health coverage