How to Choose a Thermometer 2026 | By Measuring Type, Who It's For and Hygiene
Prices, point rates and sale details in this article were checked by the mottoku editorial team against each service's official site and API (as of the last update). Always check the official sites for current prices, stock and campaign terms.
The short answer: an underarm predictive thermometer as your main one, with a no-touch or ear model in support
All thermometers look like they do the same job, but where they measure and what the number means differ quite a lot. Underarm predictive and direct-reading models show a value that isclose to your internal body temperature, while forehead no-touch and ear models show a value that isestimated from the radiant heat of the skin surface or the eardrum. So when your usual temperature looks a few tenths of a degree off after you switch models, it is not a malfunction. The two devices are simply measuring different things.
If you keep only one at home, an easy-to-read, stableunderarm predictive thermometeris the safe choice. Then, if you want to measure without waking someone at night, have a child who cannot sit still, or need to check many people in turn, addinga no-touch (forehead) or ear model as a second onemakes things much easier. Trying to cover every situation with a single thermometer usually leaves you unhappy: it is either fast but worrying, or accurate but slow.
This article is general information, not medical advice.If you have a high fever or symptoms that persist, do not rely on the thermometer reading alone; see a medical professional. Prices change over time, so please check the current listings on each online shop or in store for actual amounts.
There are only four things to look at when choosing.(1) Where you measure (underarm / ear / forehead) (2) Predictive or direct-reading (balance of speed and stability) (3) Who you mainly measure (infants or adults) (4) Hygiene (waterproofing, tip covers, whether the whole unit can be washed). Work through them in this order and the right one for your home will come out naturally.
The number means something different depending on where you measure
The first thing that trips people up is that underarm, ear and forehead readings come out slightly different. This is not a question of which is more accurate; the measured spot is different. Keeping the general tendencies in mind means you will not panic when you change models.
| Where you measure | Type | Measuring speed | Quirks and cautions |
|---|---|---|---|
| Underarm (axilla) | Predictive / direct-reading (electronic) | Predictive: 15 seconds to 1 minute; direct-reading: 5 to 10 minutes | The household standard. Placement and skin contact change the result |
| Ear (eardrum) | Infrared ear type | 1 to a few seconds | Quick, but the angle in the ear canal and earwax can throw it off |
| Forehead / temple | Infrared no-touch type | 1 to 2 seconds | Hygienic with no contact, but strongly affected by outdoor temperature, sweat and distance |
What to remember is thatmeasuring the same person at the same moment underarm and on the forehead normally gives different numbers. That is exactly why it matters most to measure your own usual temperature with the thermometer you will actually use, in the same way. For example, if your usual reading on a no-touch model is 36.2℃, then a reading above 37.2℃ on that model tells you that you are 1℃ higher than usual. Think of the baseline as something you build yourself for each device.
Predictive vs. direct-reading: how the fast number is produced
Underarm electronic thermometers come in two types that arrive at the display differently. Understanding this explains the complaints you often see in reviews, such as the predictive one is off every time or the direct-reading one takes too long.
A predictive model reads ahead from the rate of rise
A predictive model calculates internally how the temperature rises at the start (the slope of the curve) and shows ahead of time, in about 15 seconds to 1 minute, the value it expects you would settle at if you left it in for 10 minutes. It is fast and accurate enough for everyday use and is the mainstream choice for homes today. However,if contact is loose at the start, the rise is read as slower and the result can come out low. A typical case is right after placing it against a cold underarm.
A direct-reading model waits until the temperature settles
A direct-reading model does not predict; it takes 5 to 10 minutes until the actual temperature is stable. It takes time, but as long as it is placed correctly, the reading is stable. Many models aretwo-in-one predictive plus direct-reading types that switch to direct-reading mode if you leave them in place after the predicted value appears, so you can use prediction on a rushed morning and direct-reading on a night when you want to be sure, all with one device.
If a predictive model seems to give a different value every time, the cause is more oftenvariation in how it is placedthan a faulty unit. Wipe off sweat, put the tipin the center of the armpit hollow, pointing up from below at an angle, and press your arm against your body to close it. Doing just these three things the same way every time cuts the variation considerably.
No-touch and ear models trade speed for some error
No-touch and ear models get a lot of attention during infection season. Being able to measure in a few seconds without touching is a big attraction, but they also come with the weakness thatthey read radiant heat from the surface or eardrum by infrared, so they are easily affected by the environment and technique. To avoid thinking after you buy that it is less reliable than you expected, learn their quirks first.
Where no-touch (forehead) models trip you up
- Pulled around by outdoor temperature: Right after coming in from the cold outdoors the forehead is chilled and the reading tends to be low. Conversely, in direct sunlight or near a heater vent it tends to be high. The basic rule is to let it adjust to room temperature for 5 to 10 minutes before measuring
- Distance and angle: Each model has a proper distance of so many cm from the forehead. Holding it too close, too far, or at a slant makes the value fluctuate. Following the distance recommended in the manual makes more difference than you would expect
- Sweat, bangs and makeup: A wet forehead or bangs in the way disturb the radiant heat. Bare a dry forehead before measuring
Where ear models trip you up
An ear model reads eardrum temperature, so it is close to core body temperature, butif the probe is not aimed straight toward the back of the ear canal, it picks up the temperature of the ear wall along the way and reads low. For children, gently pull the ear back and up to straighten the canal before inserting; this keeps the reading steady. A lot of earwax can block the way, so it is reassuring to measure twice in a row and use the values that are close.
You will get along with no-touch and ear models if you accept that they are not tools for getting an exact absolute value but screening tools for quickly catching a change from your usual temperature. If a suspicious high reading appears, make the final check by retaking it underarm with predictive plus direct-reading. This two-step approach is the realistic one for a home.
The best answer changes with who you measure: the right pick by type
Even if you just want a thermometer, the right answer is quite different depending on who you mainly measure. Match it to your family makeup.
| Household or situation | Main thermometer | Handy second one to add |
|---|---|---|
| Adults only, health management is the main use | Underarm predictive plus direct-reading | None needed (one is enough) |
| Infant, to measure after putting the baby to sleep | No-touch (forehead) | Underarm predictive for confirmation |
| Toddler who cannot sit still | Ear or no-touch | Underarm predictive to double-check |
| Looking after an elderly family member | Underarm predictive with a large display and backlight | No-touch for quick checks on rounds |
| Family members get sick at the same time | Underarm predictive | One more (to avoid fighting over it and to take turns disinfecting) |
The key point is thatif you make a no-touch or ear model your main one, keep an underarm model nearby for confirmation. When a screening reading comes out high, having a partner you can calmly retake it with keeps your judgment steady. On the other hand, an adult-centered household that starts with a no-touch model tends to get thrown off by the error and end up buying an underarm one anyway, which is a roundabout route.
Living alone, shared, or used a few times a year: how often you use it changes the right thermometer
The right answer depends not only on who measures but also on what kind of life it is used in and how often. Even with the same underarm predictive type, someone who measures every morning and someone who takes it out a few times a year should care about very different things.
If you live alone and are the only user
When you are the only one using it, the hygiene hurdle drops a lot. You do not have to worry about tip covers or disinfecting between users, which means you canput everything into accuracy and ease of use. An underarm predictive model with a large display and an easy-to-hear beep is recommended. When you have a fever your head is foggy, and squinting at a tiny screen is hard.
What to avoid is making a no-touch thermometer your main one. When you live alone, you have to point it at your own forehead, and keeping the distance and angle yourself is surprisingly hard, and the numbers are hard to read in a mirror. The worse you feel, the sloppier you get, and readings near your usual temperature will scatter. The realistic role for a no-touch model is as a backup for when you do not have the energy to measure underarm.
If a family shares it
For shared use, hygiene and turnaround come first. There are three key points.The body is waterproof (the whole unit can be washed)、Measuring is quick、The case can stand upright. On a morning when the family takes turns, whether each person is done in ten-odd seconds or takes several tens of seconds makes a big difference in how it feels.
Also, the larger the family, the better it works to own several rather than pass one around. Keep one for the person with a fever and have the healthy family members use another. If you have two of the same model, your usual-temperature records are taken under the same conditions and are easy to compare. On the other hand, mixing two thermometers that differ in maker and type leaves you unable to tell whether it is higher than yesterday or just a different thermometer.
If it is rarely used, taken out only a few times a year
This is the pattern that fails most easily. On the day you feel feverish, you take it out and the battery is dead, or the display is too faint to read. A thermometer spends nearly all its time in a drawer, so the question iswhether it can withstand long periods of sitting unused.
When choosing, look for one whose battery is a common button cell and that clearly states you can replace it yourself. Single-use types with no battery replacement tend to need replacing before their life is up, especially in households that use them rarely. Also, the more features a model has, the harder it is to judge standby current drain and display deterioration, so if you use it rarely, it is actually more reliable to keep the features simple.
If storage space is limited
If it has to fit on a bathroom shelf or in a corner of the medicine box, think aboutthe dimensions including the caserather than the size of the thermometer alone. Pen-shaped ones are long and thin and slip into gaps, but ear models have a thick measuring part and a box-shaped case that takes up room. Some no-touch types have a gun-like grip, and lying one down in a drawer takes up surprising space.
If you are unsure where to put it, use this test: can a person with a fever reach it from their bed? If it is tucked away at the back of the medicine box, you end up just waiting and seeing without taking a temperature.
About 20 seconds and ±0.1℃: reading the numbers and marks on a product page
A thermometer spec sheet has few items, so each number carries a lot of meaning. Misreading them causes gaps such as slower than expected or the display differing slightly.
The about X seconds for measuring time is the shortest value, not a guarantee
Notations like about 20 seconds or about 15 seconds on predictive models area guide for good conditions. If the underarm is cold or the thermometer is placed too shallow, some models cannot settle on a prediction, so the time stretches or the prediction is dropped and it moves to direct-reading. A catalog figure differing from your experience is not a malfunction; it is because a predictive model by design reads ahead from interim progress. When comparing, rather than lining up the seconds of several models, it is more practical to check in the spec section how it behaves when the prediction misses (whether it automatically switches to direct-reading).
±0.1℃ is the display step, not accuracy; accuracy is a separate item
This is easy to confuse.Display unit(such as 0.1℃ steps) andinstrument errorormeasurement accuracyare different things. A thermometer as a medical device has a permitted tolerance within a set temperature range, and a spec sheet may write it with a range, such as ±0.1℃ (35.0 to 38.0℃). Some models are written so that the tolerance widens in the high or low ranges outside that range, so if you want to look closely at readings during a high fever, reading how the range is written will give you more confidence.
The medical device certification number and the controlled medical device mark
Products sold in Japan as thermometers carrya medical device certification number. It is often written as a string of digits and letters in the spec section of the product page or at the bottom of the package image. A product that lacks it and is described only as a thermometer or no-touch thermometer may not be intended for measuring body temperature. A tool that measures the surface temperature of objects is different from one certified as a clinical thermometer, so this is a mark you should check before buying.
Other marks that are easy to overlook
| Waterproof ratings such as IPX7 | The higher the number, the better it resists water. An IPX7-equivalent unit is designed to withstand brief immersion, and washing the whole thing under running water is easy. A model that is only splash-resistant will break if soaked. |
| Backlight / illuminated display | Directly affects how readable it is when you measure in a dark bedroom. For households that measure often at night, it is close to an essential feature. |
| Memory function (X readings) | A function that keeps past readings on the unit. Some keep only one reading, while others keep multiple readings or readings for several people. It helps when you want to follow progress. |
| Auto power-off | Prevents battery drain if you forget to switch it off. For households that use it rarely, whether it has this decides how long the battery lasts. |
| Battery type (LR41 / CR2032, etc.) | Information that determines whether it can be replaced and how easy it is to find. If the type is not stated, it may not have been built to be replaced. |
For no-touch types, the measuring distance notation also matters. Premises differ by model, such as 1 to 3 cm or 3 to 5 cm, and missing that distance tends to give a lower reading. Treat the distance in the manual as a condition, not a rough guide.
Check in this order before pressing the order button
With a thermometer, even if you notice after delivery that it is not what you expected, it is a hygiene product, so you tend to end up using it anyway. That is why the checking should be done before purchase. Going through the list from the top removes the fatal mismatches first.
- Is it certified as a thermometer?First look for the medical device certification number. A product without it may be a device that measures something other than body temperature. If it is off, the number you get cannot be treated as body temperature in the first place and is no basis for deciding whether to see a doctor.
- Does the measuring spot suit how your household uses it?Check which of underarm, ear and forehead it is, and whether it suits the ages in your family. If this is off, you will keep feeling the stress of not being able to measure well every time, and end up buying another one.
- Predictive or direct-reading, and can you switch?Check whether it is predictive only, or whether it can move straight on to direct-reading after the prediction. Without switching, you have no way to verify when you want to retake your usual temperature precisely or when the predicted value leaves you doubtful.
- Waterproof or splash-resistant?If you share it, whether it can be washed whole is the dividing line. Washing a splash-resistant model lets water inside and can garble the display or corrode the battery contacts. Read the rating in the spec section too.
- Battery type and whether it can be replacedIf it can be replaced, note the type and see whether it is a standard that is easy to get. Skip this and you get the worst combination: on the day you need it the battery is dead, and the matching battery is not available right away.
- Display size and backlightEnlarge the sample display on the product image. If the digits are small, you will misread them when you are fuzzy with fever or in a dark bedroom. Reading 38.5 as 36.5 is no laughing matter.
- Does it need tip covers, and how easy are the consumables to get?Some ear models require probe covers. If the covers are a proprietary item with a model number that is hard to find, the thermometer becomes unusable as soon as you run out. Check whether it can be used without covers and whether a substitute works.
- Case shape and storageCan it stand upright, and does it fit in its fixed place in the bathroom? If the case is too big, it gets pushed to the back of a drawer and does not come out when you need it.
- Beep volume and whether it can be silencedFor households that want to measure at night without waking the child, having a mute or volume setting matters. On the other hand, if an elderly family member will use it, a model with too quiet a beep means they will not notice that it has finished.
- Warranty, repair and calibrationThermometer readings can drift over the years. Checking whether the maker has a contact point for accuracy checks brings more peace of mind when you use it for a long time.
When reading reviews, rather than impressions like fast or accurate, pick out the ones that saywhich spot it was used on, on whom, and in what environment. A review of an ear model for an infant and a review of an underarm model for an adult mean completely different things even with the same star rating.
Shape, environment and storage: physical checks so you are not in trouble after it arrives
A thermometer is a small tool, so people tend to think the size does not matter, but in fact quite a few people stumble over shape and environmental conditions. Here are the physical conditions to settle before buying.
What tip shape and thickness mean depends on who you measure
Underarm thermometers broadly come instraightandangled-tiptypes. The angled type follows the armpit hollow easily and makes it easier to find the angle when measuring a sleeping child or an elderly person. The straight type, on the other hand, is easy to handle when measuring yourself and easy to wash. Neither is better; it is a matter ofwho is measuring whom changing which is easier to hold.
For ear models, the thickness of the probe (the part that goes into the ear) is decisive. An infant's ear canal is narrow, and a thick adult probe cannot reach deep and ends up picking up the temperature of the wall near the opening instead of the eardrum. If you want to use it on an infant too, be sure to check the stated age range.
Operating environment temperature, an often overlooked condition
Most thermometers have a specifiedoperating environment temperature(for example, 10 to 40℃). This is the condition for the thermometer itself to work correctly, and outside this range the display cannot be trusted. In winter, a thermometer left in an unheated hallway or entryway and then used as is may be chilled through, and its prediction may be unstable.Just leaving it in a room-temperature place for a few minutes before useavoids this error.
No-touch types are affected even more. They read surface temperature by infrared, so the larger the temperature gap between the unit and the target, the easier it is to be off. For households that use a model at the entryway, keeping the unit indoors is the first step to ensuring accuracy.
Storage conditions, and what happens in long-term storage
| Hot places | Inside a car, a sunny window sill, near heaters. Resin deforms, the display deteriorates and the battery drains faster. |
| Humid places | Near the sink in a bathroom is convenient, but if you keep exposing a splash-resistant model to humidity, the battery contacts get damaged. Store it in its case. |
| Back of the medicine box | There is little physical damage, but you lose the chance to check the remaining battery. Get in the habit of checking once a year. |
| Places children can reach | Button batteries can cause accidental swallowing. Check that the battery cover is screwed on or hard to open. |
Checking compatibility when choosing an app-linked type
Some thermometers link to a smartphone to keep records. In that case, what to check isthe supported OS versionandthe connection method. Bluetooth models may not connect if the smartphone OS is old, and if the app is discontinued, the recording function stops working. Treat the linking feature as only an extra, and choose based onwhether it is complete as a thermometer even without the appso you will not be in trouble later.
We recommend trying how to open the battery cover once on the day you buy it. If it is a screw type, you will also learn what screwdriver size is needed, and you will not have to hunt for tools on the day the battery runs out.
Measuring correctly matters more than features: technique and building your usual temperature
Even a high-end thermometer gives unreliable numbers if you measure carelessly. Conversely, even a cheap predictive model is useful enough if you keep your technique consistent. Before spending money, get this part right first.
- Wipe off sweat before placing itA wet underarm or forehead reads low because of evaporative cooling. Measuring on dry skin is the basic premise. Wait about 30 minutes after a bath, exercise or a meal.
- Underarm: push up from below for close contactPlace the tip diagonally from below in the center of the hollow and close your arm snugly against your body. Moving throws off the prediction, so hold a child in your arms to keep them still.
- No-touch: let it adjust to room temperatureAvoid measuring right after coming from a cold or hot place; wait 5 to 10 minutes first. Keep to the distance recommended in the manual.
- Ear: straighten the ear canalGently pull the ear back and up and aim the probe deep. Measuring twice in a row and using the close values keeps it stable.
- Measure your usual temperature when you are wellMeasure morning, noon and night to learn your own baseline for that device and that technique. This becomes the yardstick for judging a fever.
Usual temperature varies from person to person, and even in the same person it changes over the day, lower in the morning and higher in the late afternoon. A reading of 37.5℃ or higher is commonly taken as a guide for fever, but for someone whose usual temperature is low, even 37.0℃ can be much higher than usual.Looking not only at the absolute value but at how much it has risen from your usual temperaturelets you catch changes in your condition early.
Hygiene in infection season: washing, tip covers and how to rotate
Households that share a single thermometer are the most likely to regret how it handles hygiene. A thermometer used on someone who is unwell should not go straight to the next person. Checking the following points before you buy gives peace of mind.
- Waterproof and washable: If it is a type where not just the tip but the whole body can be washed, you can wash it after use and wipe it with alcohol. In infection season this difference is big
- Disposable tip covers (probe covers): Some ear and underarm models let you change the cover for every measurement. This makes it easier to keep things clean when measuring family members in turn
- The advantage of no-touch models: They do not touch the skin at all, so it is easy to avoid contact from person to person. When measuring many people in turn, they have the edge in hygiene
If you share one, do not skimp on the effort: wipe the tip with a clean cloth or alcohol wipe after every measurement, and change the cover for each person. On ear and no-touch models, oil from hands or dirt on the sensor lowers accuracy, so keeping the lens and sensor clean is not only about hygiene but also about accuracy.
Ways to buy that suit a thermometer: batteries, a spare and seasonal demand
A thermometer is not an expensive appliance, but it is a tool that causes real trouble if it will not work on the night a fever comes. We recommend buying with priority onbeing able to use it reliably when neededrather than on a low price.
Check the battery type first
Many thermometers run on button (coin) batteries or AAA batteries.Button battery types differ by model, and when one runs out the batteries you have at home may not fit, so note the type when you buy and keep one spare for peace of mind. Replace the battery early once the low-battery indicator appears. Some no-touch models drain their batteries a little faster.
Prepare outside the peak season
Thermometers, especially no-touch ones, tend tosee demand concentrate in winter to early spring when infections spread, with stock running short and prices rising. Rather than hunting in a panic at the peak of the outbreak, choosing calmly when demand has settled tends to give you better options and better condition.
If you want to buy a little more advantageously,stacking drugstore point days, Rakuten Shopping Marathon and each shop's sale periodsis the standard way. Thermometers are sometimes treated as daily necessities and can qualify for bulk-buy campaigns. However, the point rates and conditions change with the season, sobe sure to check the specific rates and caps on each official page before buying. Decide on the model first, then compare price and points, and your judgment will stay steady.
Another way of buying that suits a thermometer is a spare one. When family members get sick at the same time, one does not go around, and you end up waiting for it to dry after each disinfection. For households with infants or elderly members, or where you measure several times a day while caring for someone, setting up two from the start (underarm predictive plus no-touch or ear) resolves the fighting over it and the hygiene worries all at once.
Frequently asked questions
My usual temperature seems to have changed after I switched models. Is it broken?
Most of the time it is not a malfunction. Underarm, ear and forehead measure different spots, so the displayed value differs even for the same person. When you start using a new thermometer, measure your usual temperature again while healthy and rebuild your baseline for that model. What matters is not the absolute value but how much it has risen from your usual temperature.
Predictive or direct-reading: which should I buy?
For everyday use, a predictive model that measures in a few tens of seconds is enough. If there are also times you want to be more certain, a predictive plus direct-reading model that switches to direct-reading after the predicted display is handy. You can use prediction on a rushed morning and direct-reading on a night when you want to take your time, all on one device. Direct-reading-only models take 5 to 10 minutes, so for homes a switchable model is easier to use.
Can I trust a no-touch thermometer?
It is quick and hygienic, but it is a type that is prone to error from outdoor temperature, sweat, measuring distance, and a bath or exercise just before. The premise is to let it adjust to room temperature and follow the recommended distance. Think of it as a screening tool for quickly catching changes from your usual temperature rather than a tool for absolute values, and if a suspicious value appears, retaking it underarm gives peace of mind.
Which type is best for babies and toddlers?
For children who cannot sit still, no-touch (forehead) or ear models that measure in a few seconds without touching are suitable. The trick with an ear model is to pull the ear back and up to straighten the canal before inserting. When you want to confirm more accurately, you can also hold the child and use an underarm predictive model. Making a no-touch or ear model the main one while keeping an underarm model nearby for confirmation is the realistic setup.
Why do the values vary each time I measure?
Body temperature varies with time of day, exercise and meals, and variation in technique has a large effect. Underarm models swing with placement (contact and angle), and no-touch models with distance and outdoor temperature. Using the same thermometer and the same technique each time, and wiping off sweat before measuring, reduces the variation. If it is still extremely unstable, suspect the battery or a malfunction as well.
At what temperature should I consider it a fever?
A reading of 37.5℃ or higher is commonly used as a guide for fever, but usual temperature varies from person to person. For someone whose usual temperature is low, even 37.0℃ can be quite high, so it is important to know your own usual temperature. If you have a high fever or symptoms that persist, do not decide on your own from the number alone; see a medical professional.
How can I avoid being stuck when the battery dies?
Battery types differ by model, so note the type when you buy and keep one spare for peace of mind. Button batteries often do not match the ones you have at home. Replace the battery early once the low-battery indicator appears, and if the display looks wrong, recheck the battery and your technique. If it still acts up, it may be a malfunction.
How many thermometers should a home have?
For an adult-centered household, one underarm predictive model is enough for daily use. Households with infants or elderly members, where you measure many times while caring for someone, or where family members get sick at the same time, will feel safer with two: underarm predictive plus no-touch or ear. This reduces the stress of disinfecting and fighting over it. Stock tends to run short in outbreak season, so households that need one are wise to prepare when demand has settled.
About how many years should I use a thermometer before replacing it?
There is no clearly set expiry period, but the sensor and electronic parts of an electronic thermometer deteriorate over time, so many households review it at around 5 years. If you have dropped it, submerged it, the display is missing segments, or it now shows a different tendency from before under the same conditions, consider replacing it regardless of its age.
I measured at the same time, but two thermometers show different numbers. Which one is right?
It does not necessarily mean one is broken. Differences of about 0.1 to 0.2℃ can arise from the type, the measuring spot, and even within the instrument error range. What matters is to choose one as your reference and record your usual temperature with that same one. If you fix the reference, you can correctly judge whether it is higher than usual even if the absolute value is slightly off.
Can I wash a thermometer with water?
Only models marked waterproof can be washed with water. If the spec section lists a waterproof rating such as IPX7, you can wash it under running water, but soaking a splash-resistant model lets water inside and causes failure. If it is not waterproof, the basic method is to wipe the tip and body with a cloth moistened with disinfectant ethanol. In either case, avoid hot water and boiling.
The battery had run out without my noticing and I could not use it when I needed it. Is there a way to prevent this?
The sure way is to choose a model whose battery can be replaced and keep one battery of the matching type together with the thermometer's case. It also helps to make a habit of switching it on once a year, such as at a change of season or the start of the fiscal year, and checking that the display has no missing or faded segments, so you are not caught off guard on the day you need it.
* This article is for reference only. Sale dates, point rates and campaign terms can change, so check each store's official page for the latest details.