Touching someone’s forehead with the back of your hand is a surprisingly decent screening tool, but it catches fever only about 87–89 percent of the time and falsely flags a healthy person as feverish in roughly half of all checks. A thermometer remains the only way to get an actual number. How you use one, which type you pick, and where you measure all change the reading you get, sometimes by a degree or more. Knowing what each method can and cannot tell you makes the difference between catching a real fever and either missing one or panicking over nothing.
Your Baseline Is Probably Not 98.6 °F
Before you can decide whether a reading is high, you need to know what “normal” looks like. The famous 98.6 °F figure comes from a 19th-century German physician named Carl Wunderlich, whose work became the foundation of clinical thermometry.1PubMed Central. Thermometry and interpretation of body temperature Modern data tells a slightly different story. A large-scale analysis of body-temperature measurements found the overall mean to be about 98.1 °F (36.7 °C), with a standard deviation of roughly 1.1 °F.2PubMed. The daily, weekly, and seasonal cycles of body temperature analyzed at large scale That means plenty of healthy people walk around at 97.5 °F or 98.8 °F and are perfectly fine.
Your temperature also shifts throughout the day, bottoming out in the early morning (around 6:00–8:00 a.m.) and peaking in the evening (around 6:00–8:00 p.m.).2PubMed. The daily, weekly, and seasonal cycles of body temperature analyzed at large scale A reading of 99.0 °F at 7 a.m. is more notable than the same number at 7 p.m. If you regularly check your temperature when you feel healthy, you get a personal baseline that makes future readings easier to interpret.
How Thermometer Type and Body Site Change the Number
Not all thermometer readings are interchangeable. Rectal temperature is generally treated as the closest practical proxy for true core body temperature. Every other site tends to read lower, but the gap is not a neat, consistent offset you can just add or subtract.
In one emergency-department study, the average oral temperature was about 1.1 °F lower than rectal, but in individual patients the oral reading could be as much as 2.9 °F below the rectal or 0.7 °F above it.3PubMed Central. Oral and Tympanic Membrane Temperatures Are Inaccurate to Identify Fever in Emergency Department Adults That is an enormous range. A tympanic (ear) thermometer did no better: although the average ear reading was only about 0.2 °F off from rectal, any single ear measurement could swing from 1.6 °F below to 2.1 °F above the rectal value.3PubMed Central. Oral and Tympanic Membrane Temperatures Are Inaccurate to Identify Fever in Emergency Department Adults
A separate comparison of digital thermometers against standard mercury rectal readings concluded that electronic rectal measurements came closest, while electronic tympanic, oral, and axillary (armpit) measurements all showed unacceptable variability.4PubMed. Accuracy of Digital Tympanic, Oral, Axillary, and Rectal Thermometers Compared with Standard Rectal Mercury Thermometers That does not mean you should never use an oral or ear thermometer at home. It means you should treat any single peripheral reading as an estimate, not a verdict. If the number sits right on the borderline of fever, re-check rather than making a decision based on one measurement.
Practical Tips for More Reliable Readings
- Oral: Keep your mouth closed for at least a minute before inserting the thermometer. Drinking hot or cold liquids, breathing through your mouth, or eating recently can throw the reading off.
- Ear: Pull the ear gently up and back to straighten the canal in adults, and aim the sensor toward the eardrum. Earwax buildup and small ear canals in infants reduce accuracy.
- Armpit: Press the arm snugly against the body and wait the full time the device specifies. Armpit readings tend to run lower than oral by about half a degree Fahrenheit, and they are the least reliable of the common sites.
- Rectal: Most accurate for infants and young children. Use a flexible-tip thermometer, insert no more than about an inch, and hold the child still.
Checking for Fever Without a Thermometer
When no thermometer is available, touch is what most people reach for. You press your lips or the back of your hand to the person’s forehead and make a judgment call. Research suggests this works better than you might expect for ruling fever out, but it is unreliable for confirming fever.
A systematic review and meta-analysis of caregiver touch in children found a pooled sensitivity of about 87.5 percent, meaning it catches most true fevers, but a specificity of only about 55 percent, meaning it falsely labels many non-febrile children as having a fever.5PubMed. Accuracy of Tactile Assessment of Fever in Children by Caregivers: A Systematic Review and Meta-analysis An earlier meta-analysis of maternal touch specifically reported even higher sensitivity at about 89 percent but roughly the same specificity around 50 percent, leading the authors to note that touch is “perhaps more useful to exclude fever rather than to ‘rule in’ fever.”6Journal of Tropical Pediatrics. The Accuracy of Mother’s Touch to Detect Fever in Children: A Systematic Review
In practical terms, if you touch a child and they feel completely normal temperature-wise, you can have reasonable confidence they probably do not have a significant fever. But if they feel warm, you cannot be sure. Anxiety and expectation tend to push parents toward calling a child feverish when they are not. One study found that medical staff had notably better specificity than caregivers, likely because the caregiver’s emotional state biases their judgment.7PubMed. Reliability of perception of fever by touch
Where to Touch
Most people go straight for the forehead, and it is a reasonable spot. But research indicates the abdomen and neck perform at least as well and possibly better for detecting warmth by hand.7PubMed. Reliability of perception of fever by touch The torso stays closer to core temperature than the forehead, which can be cooled by air exposure or sweat. If you are trying to check yourself, placing a hand on your own abdomen and comparing it to your sense of what feels normal can be informative, though self-assessment is inherently less accurate than having someone else check you.
Are Fathers as Good as Mothers at This?
A question that has actually been studied: one cross-sectional trial in children found that fathers and mothers detected fever by touch at nearly identical sensitivity rates, around 86–88 percent. Fathers were somewhat less specific, meaning they were slightly more likely to over-diagnose fever, but the practical difference was small.8PubMed Central. A Cross-Sectional Study on Subjective Fever Assessment in Children by Palpation: Are Fathers as Reliable as Mothers? Either parent’s hand does a similar job.
Other Physical Signs That Suggest Fever
Temperature is not the only thing that changes when your body mounts a febrile response. Several other observable signs can give you clues when a thermometer is unavailable.
Heart rate rises predictably with temperature. A large study involving children attending urgent and emergency care found that for every 1 °C increase in body temperature, heart rate increased by about 12 beats per minute on average, with the effect being larger in younger children and smaller in teenagers.9PubMed Central. The association between temperature, heart rate, and respiratory rate in children aged under 16 years attending urgent and emergency care settings In adults, a separate study found the rise to be about 8.5 beats per minute for each degree Celsius.10PubMed. Fever and cardiac rhythm So if you or someone else has a resting pulse that is noticeably faster than usual and other symptoms like body aches or chills, a fever is a reasonable guess.
Chills and shivering are your body’s way of generating heat to raise its thermostat set point. They typically occur during the rising phase of a fever, as your muscles contract rapidly to produce warmth. If you are shivering under blankets and it is not cold in the room, your body may be working toward a higher temperature.
Flushed skin, especially across the cheeks and chest, reflects increased blood flow to the skin surface during fever. Glassy or watery eyes, general fatigue, and loss of appetite are all common accompaniments. None of these alone confirms a fever, but several together make a strong circumstantial case.
Non-Contact Infrared Thermometers and Their Limits
The forehead-scanning thermometers that became ubiquitous during the COVID-19 pandemic seem ideal: fast, no-touch, and easy to use. Their accuracy, however, falls well short of what most people assume.
A clinical evaluation of multiple non-contact infrared thermometer (NCIT) models found that their bias ranged from under-reading by about 0.9 °C to over-reading by about 0.2 °C. Individual measurements could deviate by as much as 3 °C below or 2 °C above the reference temperature. Depending on the model, between 48 and 88 percent of readings fell outside the accuracy range the manufacturer claimed on the label. When it came to actually detecting a temperature above 38 °C, the sensitivity of the various models ranged from zero to 0.69.11Nature. Clinical evaluation of non-contact infrared thermometers A sensitivity of zero means the device missed every single fever in the test sample. A sensitivity of 0.69 means it caught about two-thirds of them. Neither is reassuring.
These devices are heavily influenced by ambient temperature, distance from the skin, sweat on the forehead, and whether the person has been outside in cold or hot weather. They work best in climate-controlled indoor environments when used at a consistent distance on dry skin. If you rely on one at home, treat it as a screening tool rather than a definitive measurement. A normal reading from a forehead scanner does not guarantee the absence of fever, and a borderline reading warrants a follow-up with an oral or rectal thermometer.
Wearable Devices and Continuous Monitoring
Smartwatches and fitness bands with skin-temperature sensors have opened up the possibility of tracking body temperature around the clock. Rather than catching a single snapshot, these devices record patterns over days or weeks, which gives them an interesting advantage: they can detect a rise relative to your personal baseline rather than comparing against a population average.
A feasibility study on continuous fever monitoring with wearables found a significant increase in skin temperature during periods when participants reported symptoms, averaging about 0.63 °C above their own baseline. The researchers noted that the variability between individuals was substantial enough that using a fixed fever cutoff like 38 °C across all users would not be appropriate.12Scientific Reports. Feasibility of continuous fever monitoring using wearable devices In other words, wearables may eventually be better at noticing that your temperature has jumped compared to your own normal, even if the absolute number never crosses the traditional fever line.
The technology is still early. Most consumer wearables measure skin temperature at the wrist, which is influenced by movement, ambient air, and how tightly the band sits. They are not yet reliable enough to replace a thermometer for clinical decisions. But for picking up trends, like noticing that your temperature climbed steadily overnight, they provide a kind of information that a single morning thermometer check cannot.
Why Fever Can Be Easy to Miss in Older Adults
Age changes the rules. Older adults tend to run cooler at baseline and are less able to mount the same vigorous fever response as younger people.13PubMed Central. Altered Febrile Responses in Older Adults: A Systematic Review This means a temperature of 99.5 °F in a 78-year-old could represent a more significant infection than 101 °F in a 30-year-old.
Research on infections in elderly patients found that the fever response may be absent or blunted roughly 20–30 percent of the time.14PubMed. Fever in the elderly That is a real problem, because fever is often the first signal that something is wrong. When it does not show up, diagnosis gets delayed, and outcomes can suffer. For an older adult, a sudden change in mental clarity, new confusion, unexplained fatigue, or loss of appetite may be the equivalent of a fever spike in a younger person. Caregivers who rely solely on a thermometer reading to decide whether an older family member is sick can miss serious infections.
For this population, knowing their personal baseline temperature matters even more than for younger adults. If someone normally runs at 97.2 °F, a reading of 99.0 °F represents nearly a two-degree jump and warrants attention, even though most fever guidelines would call 99.0 °F normal.
Temperature in Heat-Related Illness
Checking temperature takes on a different urgency when the concern is not infection but environmental heat exposure. Heat exhaustion and heat stroke are distinct conditions with different temperature profiles. Heat exhaustion typically involves dizziness, nausea, heavy sweating, and mildly elevated body temperature. Heat stroke, the dangerous one, is defined by a core temperature above 104 °F (40 °C) along with changes in mental status like confusion, slurred speech, or loss of consciousness.15PubMed. Heat-related illness in athletes
This is a context where thermometer accuracy genuinely matters. A forehead scanner on a sweating athlete outdoors in August is going to give you garbage data because of evaporative cooling on the skin. Oral readings can be thrown off by heavy breathing. In athletic and military settings, rectal temperature is the accepted standard for assessing heat stroke because it reflects core temperature most reliably. At home, if someone collapses after prolonged heat exposure and seems confused, do not wait for a perfect temperature reading. Call emergency services. The mental-status change is the red flag, and it is observable without any device at all.
Recognizing Cold-Related Temperature Drops
On the opposite end, hypothermia (core body temperature below 95 °F / 35 °C) presents its own detection challenges. Many home thermometers do not read below about 94 °F, which means they will either display an error or give an artificially high reading in someone who is genuinely hypothermic.
Without a thermometer, the signs of hypothermia progress in a fairly predictable sequence. Mild hypothermia brings shivering, clumsiness, and confusion. As it worsens, shivering stops (which is paradoxically a bad sign, because the body has exhausted its ability to generate heat), speech becomes slurred, and the person becomes increasingly drowsy. In severe cases, core temperature can drop below 75 °F, and people have survived with vital signs so minimal they appeared dead.16Mary Ann Liebert, Inc., publishers. Deep accidental hypothermia with core temperature below 24°c presenting with vital signs The clinical rule in emergency medicine is that a hypothermic patient is not declared dead until they are warm and dead.
For practical purposes, if someone has been exposed to cold water or cold weather and is shivering uncontrollably, confused, or struggling to use their hands, assume hypothermia and focus on rewarming. Trying to get a thermometer reading at that point is far less important than getting them to shelter, removing wet clothing, and sharing body heat or applying warm blankets.
Capillary Refill and Peripheral Perfusion
Healthcare workers sometimes use a quick bedside test that is easy to replicate at home, though it tells you about circulation rather than body temperature directly. Press firmly on a fingernail or toenail for a few seconds until the nail bed turns white, then release. In a well-hydrated person with good circulation, the color should return to pink in under two seconds. Slower refill can suggest dehydration, poor blood flow, or shock, all of which can accompany serious illness.
Capillary refill time and the feeling of cool limbs compared to a warm trunk are used in clinical settings as quick markers of how well blood is reaching the extremities.17Archives of Disease in Childhood. Evaluating intravascular hypovolaemia by manually assessing peripheral limb coolness compared with the capillary refill time in children aged <5 years in a tropical environment: a prospective cohort study of diagnostic accuracy This is not a fever check per se. But when you are assessing a sick person without tools, cold hands and feet paired with a hot trunk can point to fever with vasoconstriction, while universally warm and flushed skin alongside a fast pulse suggests the fever is further along and the body is actively trying to dump heat.
A Brief History of Measuring Body Heat
Humans have been using touch to judge body temperature for thousands of years. Physicians in ancient Greece and medieval Arabia routinely assessed fever by hand before any measuring instrument existed. The thermometer only entered the picture in the 16th century, and it was not until the late 1800s that Wunderlich’s systematic study of thousands of patients established the temperature ranges that clinicians still reference today.18PubMed. The historical development of thermometry and thermal imaging in medicine His landmark figure of 37 °C (98.6 °F) as the “normal” body temperature persisted for well over a century before modern studies revised it downward.
Mercury glass thermometers dominated for most of the 20th century. They were accurate but slow, fragile, and posed a toxicity risk if broken. Digital thermometers replaced them starting in the 1990s, and by now mercury thermometers have been phased out of most clinical use and are banned from sale in some jurisdictions. Ear thermometers, forehead scanners, and temporal-artery devices followed, each trading some accuracy for convenience. The trend line is clear: people want temperature checks to be faster and less invasive, and manufacturers have obliged. The cost is that every step away from a core-temperature measurement introduces more noise into the reading.