The average human body temperature is closer to 97.9°F (36.6°C) than the 98.6°F (37.0°C) figure most of us grew up with. That famous number dates to the 1860s and has held on in textbooks and thermometer packaging despite decades of evidence that it runs a bit high. A landmark 1992 study at the University of Maryland found the mean oral temperature among healthy adults was 98.2°F, and more recent large-scale analyses suggest the number has continued to drift downward since then.1JAMA. A critical appraisal of 98.6 degrees F, the upper limit of the normal body temperature, and other legacies of Carl Reinhold August Wunderlich But a single number is a rough summary of something that shifts with the time of day, your age, your sex, where you measure, and even the decade you were born in.
Where 98.6°F Came From and Why It Stuck
The 98.6°F standard traces back to Carl Reinhold August Wunderlich, a German physician who published an enormous dataset of roughly one million temperature readings from about 25,000 patients in 1868. He declared 37.0°C (98.6°F) the average and 38.0°C (100.4°F) the threshold for fever. Those numbers became clinical gospel. When researchers at the University of Maryland revisited the question with 148 healthy volunteers and modern calibrated thermometers, they found the mean oral temperature was 36.8°C (98.2°F), not 37.0°C. The upper limit of normal was 37.7°C (99.9°F), slightly below Wunderlich’s 38.0°C. Many of Wunderlich’s broader observations held up, including that temperature peaks in the late afternoon and runs slightly higher in women, but the specific numbers he anchored to were off by a few tenths of a degree.1JAMA. A critical appraisal of 98.6 degrees F, the upper limit of the normal body temperature, and other legacies of Carl Reinhold August Wunderlich
Part of the discrepancy likely reflects Wunderlich’s instruments. Mid-19th-century mercury thermometers were bulkier, slower, and less standardized than modern digital devices. Wunderlich also measured under the arm, which tends to give different readings than the mouth. But instruments alone don’t explain everything. There’s growing evidence that humans really were warmer 150 years ago.
Human Bodies Have Been Getting Cooler
A 2020 study in eLife analyzed more than 677,000 oral temperature readings from three historical cohorts spanning birth years from the early 1800s through the late 1990s. The researchers found a steady decline of about 0.03°C per decade of birth, adding up to roughly 0.59°C in men across the full time span and about 0.32°C in women over the period studied.2eLife. Decreasing human body temperature in the United States since the Industrial Revolution This wasn’t just a measurement artifact. The decline persisted after the researchers controlled for the time of day and the type of thermometer used.
The same cooling trend has appeared outside the United States. A study of the Tsimane people, an indigenous population in the Bolivian Amazon, documented a similar drop over a much shorter period, confirming that the phenomenon isn’t limited to industrialized nations.3PubMed Central. Rapidly declining body temperature in a tropical human population The leading explanation involves reduced chronic inflammation. In the 19th century, persistent infections like tuberculosis, periodontal disease, and untreated wounds were far more common, and chronic infection keeps the immune system in a low-grade inflammatory state that raises baseline temperature. Better sanitation, antibiotics, and improved living conditions over the past century and a half have lowered that background inflammation. Some researchers have also suggested that widespread climate control, meaning heated and air-conditioned homes, may reduce the metabolic work needed to maintain a stable temperature, contributing to the downward trend.
Your Temperature Changes Throughout the Day
Body temperature follows a predictable daily rhythm, lowest in the early morning and highest in the late afternoon. The 1992 Maryland study documented a low point around 6 AM and a peak between 4 and 6 PM, with the average swing being about 0.5°C (0.9°F).1JAMA. A critical appraisal of 98.6 degrees F, the upper limit of the normal body temperature, and other legacies of Carl Reinhold August Wunderlich This means a reading of 99.0°F at 5 PM and a reading of 97.8°F at 6 AM can both be perfectly normal for the same person.
This rhythm is generated by the brain’s internal clock, located in a cluster of cells in the hypothalamus. The same clock that governs your sleep-wake cycle also drives daily oscillations in metabolic heat production.4PubMed Central. Circadian rhythmicity of body temperature and metabolism Even if you stayed in bed for 24 hours with no change in activity or food intake, your temperature would still cycle up and down. External factors like meals, exercise, and ambient temperature can nudge the curve, but they don’t create it. The rhythm persists in isolation studies where volunteers live for weeks without time cues.
Where You Measure Matters More Than You’d Think
Rectal, oral, ear (tympanic), and armpit (axillary) readings don’t agree with each other, and the differences aren’t trivially small. In one emergency-department study, the mean oral temperature was 98.3°F while the mean rectal temperature was 99.4°F, a gap of more than a full degree. Individual-level disagreement was even wider: an oral reading could fall nearly 3°F below or about 0.7°F above the simultaneous rectal reading.5PubMed Central. Oral and Tympanic Membrane Temperatures Are Inaccurate to Identify Fever in Emergency Department Adults That range matters when you’re trying to decide whether someone has a fever.
Rectal readings are generally considered the closest approximation to true core temperature. One comparison found that electronic rectal measurements tracked standard mercury rectal readings with negligible deviation, while oral, ear, and armpit readings all showed unacceptable scatter.6The European Journal of Surgery. Accuracy of Digital Tympanic, Oral, Axillary, and Rectal Thermometers Compared with Standard Rectal Mercury Thermometers In older adults, the difference becomes clinically meaningful: about 15% of elderly patients who were afebrile by oral thermometry turned out to be febrile when measured rectally.7PubMed. A comparison of oral, tympanic, and rectal temperature measurement in the elderly That’s a lot of missed fevers from a simple choice of thermometer placement.
For practical purposes, if you’re using an oral thermometer at home, keep in mind that your reading will generally run about a degree Fahrenheit below a rectal reading, and that drinking cold or hot liquids within a few minutes beforehand will throw things off. Ear thermometers are convenient but variable enough that they’re best treated as a screening tool rather than a precise measurement.
Sex, Hormones, and the Monthly Cycle
Women tend to run slightly warmer than men. The Maryland data confirmed this, and a large study of more than 18,000 adults found women had a mean oral temperature of 97.5°F compared with 97.2°F for men.8The Journals of Gerontology: Series A. Is Older Colder or Colder Older? The Association of Age With Body Temperature in 18,630 Individuals The difference is modest in absolute terms but consistent across studies.
Much of this gap ties back to the menstrual cycle. After ovulation, rising progesterone pushes core body temperature up by roughly 0.3°C to 0.7°C compared with the first half of the cycle.9PubMed Central. Temperature regulation in women: Effects of the menstrual cycle This temperature shift is most obvious in early-morning readings before any physical activity, which is why basal body temperature tracking has long been used as a fertility-awareness method. The correlation between progesterone breakdown products and temperature is real but imperfect. An analysis of day-by-day readings found a moderate correlation between progesterone levels and basal temperature, with the relationship being roughly linear at low hormone levels but flattening out above a certain progesterone threshold.10PubMed. Descriptive analysis of the relationship between progesterone and basal body temperature across the menstrual cycle
Estrogen complicates the picture. It tends to lower the body’s thermoregulatory set-point, effectively working in the opposite direction from progesterone. When both hormones are present simultaneously, as happens in certain oral contraceptive formulations, the temperature-raising effect of progestin can be blunted.11PubMed. Estrogen modifies the temperature effects of progesterone So a woman on a combined hormonal contraceptive may not see the same clear post-ovulatory rise that someone with a natural cycle would.
Temperature Falls With Age
Older adults consistently run cooler than younger ones. That study of 18,630 adults found a difference of about 0.3°F between the oldest and youngest age groups after adjusting for sex and body mass.8The Journals of Gerontology: Series A. Is Older Colder or Colder Older? The Association of Age With Body Temperature in 18,630 Individuals A systematic review focused on older people found that oral temperatures in this group averaged roughly 1.2°F lower than the textbook norms, with smaller but still consistent gaps at rectal and ear sites.12PubMed. A systematic review of body temperature variations in older people
The clinical implication is that an older person can be running a meaningful fever at a number that looks completely normal by standard cutoffs. If your resting baseline is 97.0°F rather than 98.6°F, a reading of 99.5°F represents nearly two and a half degrees of elevation, even though it wouldn’t trip the conventional 100.4°F alarm. This is one of the reasons infections in elderly patients are sometimes caught late. Their thermoregulatory machinery becomes less responsive: the shivering response weakens, blood-vessel dilation becomes less efficient, and the overall range between the daily low and high tends to narrow.13PubMed. Age-dependent changes in temperature regulation – a mini review
What Counts as a Fever
Given all of this variation, the fever threshold depends on who you ask, where you measure, and what time of day it is. Harrison’s Principles of Internal Medicine, one of the standard medical references, defines fever as a morning oral temperature above 37.2°C (99.0°F) or a late-afternoon oral temperature above 37.7°C (99.9°F). The Merck Manual uses a slightly different cutoff of 37.8°C (100.0°F) orally or 38.2°C (100.8°F) rectally.14PubMed Central. Defining Fever For children, major pediatric guidelines generally define fever as a rectal temperature above 38.0°C (100.4°F).15PubMed Central. Threshold for defining fever varies with age, especially in children: A multi-site diagnostic accuracy study
The real takeaway is that a universal number misses the point. A wearable-thermometer study that tracked skin temperatures continuously found so much variation between individuals that applying a single fever threshold across everyone didn’t work well. People with naturally low baselines flagged late or not at all, while those who naturally ran warm flagged too often.16Scientific Reports. Feasibility of continuous fever monitoring using wearable devices Knowing your own typical resting temperature, measured at the same time and site over several days, gives you a much more useful reference than any textbook number.
Exercise and Heat Production
Vigorous exercise drives core temperature well above resting values. Working muscles generate enormous heat: in one lab study, the rate of heat production in the thigh muscles nearly doubled within the first three minutes of intense cycling.17PubMed Central. Heat production in human skeletal muscle at the onset of intense dynamic exercise Core temperature measured in the esophagus rose by about 0.55°C during exercise, while the temperature inside the working leg muscle jumped by 2 to 3°C depending on depth.18PubMed. Muscle temperature transients before, during, and after exercise measured using an intramuscular multisensor probe Core temperature typically stays elevated for a while after you stop, partly because residual heat continues to transfer from the muscles into the bloodstream.
In hot environments, the challenge becomes even steeper. Your body relies on sweating and increased blood flow to the skin to dump excess heat, and if those mechanisms can’t keep up, core temperature can climb into dangerous territory.19PubMed. Exercise under heat stress: thermoregulation, hydration, performance implications, and mitigation strategies This is why a temperature reading taken within an hour of a hard workout is essentially useless for evaluating whether you’re sick.
Sleep and the Nightly Cool-Down
Your body actively cools itself to fall asleep. Before sleep onset, blood vessels in the hands and feet dilate, flushing warm blood to the skin surface and radiating heat away from the core. This rise in skin temperature at the extremities, paired with a drop in core temperature, appears to be a powerful signal for the brain to initiate sleep. In one study, the degree of warming in the hands and feet was a better predictor of how quickly someone fell asleep than core temperature itself, melatonin timing, or subjective sleepiness ratings.20PubMed. Functional link between distal vasodilation and sleep-onset latency?
Core temperature continues to fall through the night, reaching its lowest point in the early-morning hours before climbing again before waking.21PubMed Central. Core body temperature changes before sleep are associated with nocturnal heart rate variability People who have trouble falling asleep sometimes show a blunted presleep temperature drop, which is part of the rationale behind practices like taking a warm bath before bed: the rapid skin cooling that follows the bath mimics the natural vasodilation pattern and can help nudge the temperature decline along.22PubMed. Body temperature and sleep
Body Weight, Seasons, and Other Modulators
You might assume that people with more body fat would run warmer, since fat is an insulator. The reality is surprisingly mixed. A Swiss population study found positive associations between body mass index and body temperature in men and postmenopausal women, with higher BMI linked to slightly higher temperatures even after adjusting for other factors.23International Journal of Obesity. Association of body temperature with obesity: The CoLaus study But controlled studies that measured 24-hour core temperature profiles directly found no meaningful difference between obese and non-obese participants.24The American Journal of Clinical Nutrition. Core Body Temperature in Human Obesity 25The American Journal of Clinical Nutrition. Adiposity and human regional body temperature The population-level association may reflect metabolic activity or heart rate rather than insulation. In any case, the effect is small enough that body weight is not a reliable predictor of what your thermometer will say.
Seasons also leave a mark. A large-scale analysis found that body temperatures ran about 0.2°C warmer in summer than winter, likely because ambient heat nudges the thermoregulatory system within its tolerance range rather than being fully compensated for.26PubMed. The daily, weekly, and seasonal cycles of body temperature analyzed at large scale People who spend extended time in hot environments also undergo heat acclimatization, a set of physiological adjustments that includes a small reduction in resting core temperature along with earlier and heavier sweating.27PubMed Central. Seasonal Heat Acclimatisation in Healthy Adults: A Systematic Review
Medications That Shift Your Thermostat
Several common medications interfere with the body’s ability to regulate temperature, which becomes especially relevant during heat waves. A systematic review found that drugs with strong anticholinergic properties raised core temperature by about 0.4°C during heat exposure, mainly by reducing sweating. Non-selective beta-blockers, adrenaline, and anti-Parkinson’s agents also produced small but measurable increases. Antidepressants and diuretics, despite theoretical concerns, did not significantly alter core temperature in the available evidence.28PubMed Central. The effect of prescription and over-the-counter medications on core temperature in adults during heat stress: a systematic review and meta-analysis
On the other end, common anti-inflammatory painkillers like ibuprofen don’t appear to lower daytime body temperature in healthy people, even though they work to bring a fever down when one is present.29Physiology & Behavior. Nonsteroidal anti-inflammatory drugs alter body temperature and suppress melatonin in humans That distinction makes pharmacological sense: these drugs block the inflammatory molecules that raise the thermostat’s set-point during infection, but if the set-point is already where it should be, there’s nothing to push down.
Why Mammals Settled on This Temperature Range
There’s a compelling evolutionary argument for why mammalian body temperature landed near 37°C rather than, say, 30°C or 42°C. Most environmental fungi can’t grow above about 30°C to 35°C, and each additional degree Celsius in the range between 30°C and 40°C excludes an additional 6% of fungal species from being able to infect a host.30The Journal of Infectious Diseases. Vertebrate Endothermy Restricts Most Fungi as Potential Pathogens Running warm essentially creates a thermal exclusion zone that keeps the vast majority of fungal pathogens at bay.
But warmth comes at a steep metabolic cost: mammals burn far more calories per day than a similarly sized reptile. A mathematical model that weighed the protective benefit of higher body temperature against the extra caloric expense found the optimum at 36.7°C, very close to the actual mammalian average.31PubMed Central. Mammalian endothermy optimally restricts fungi and metabolic costs This doesn’t mean antifungal defense is the only reason mammals are warm, but it offers a tidy explanation for why the set-point ended up where it did rather than a few degrees higher or lower. It also suggests a practical implication of fever: by pushing the body above its normal range, a fever extends the thermal exclusion zone even further, potentially slowing the growth of pathogens that can tolerate normal body heat.
How Your Brain Controls the Thermostat
The hypothalamus, a small structure at the base of the brain, acts as the body’s temperature control center. Within it, a region called the preoptic area contains neurons that respond to subtle changes in blood temperature. These neurons also receive signals from temperature sensors in the skin and spinal cord, integrating information about what’s happening both inside and outside the body to coordinate the appropriate response: shivering and blood-vessel constriction to conserve heat, or sweating and vasodilation to shed it.32PubMed. Role of the preoptic-anterior hypothalamus in thermoregulation and fever
During an infection, immune cells release signaling molecules called pyrogens that alter the firing of these temperature-sensing neurons. The result is that the brain temporarily raises its set-point, treating the current normal temperature as “too cold” and activating heat-generating responses until the body reaches a new, higher target. Research in mice has mapped specific neural circuits involved: a population of inhibitory neurons in one part of the hypothalamus normally keeps a downstream heat-generating area in check. When those inhibitory neurons are silenced, core temperature spikes to fever-like levels; when they’re activated, core temperature drops.33PubMed Central. A hypothalamic circuit that controls body temperature This circuit helps explain why fever feels like being cold even though you’re actually warming up: the brain has moved the goalposts, and until your body catches up, it perceives a deficit.
Toward Personalized Baselines
The growing recognition that there’s no single “normal” temperature has practical consequences for how temperature is monitored in clinical settings. Researchers testing a continuous wearable armpit thermometer noted that individual differences in resting temperature were large enough to make a universal fever cutoff unreliable. One person’s normal afternoon reading might look like another person’s low-grade fever.34PubMed Central. Investigation of Non-invasive Continuous Body Temperature Measurements in a Clinical Setting Using an Adhesive Axillary Thermometer (SteadyTemp®) Continuous monitoring devices that establish a personal baseline over days or weeks could flag a meaningful rise above that individual’s norm rather than relying on a population average. This approach is still mostly confined to research and a handful of clinical trials, but it represents a shift in thinking: from asking “is this number above the threshold” to asking “is this number unusual for this person.”