How many calories should I eat?
By Rick Campbell · Updated · Sourced to primary literature · Not medical advice
This is the most-asked question in nutrition, and the honest answer has two parts. The first is an estimate of what your body currently costs to run. The second is a decision about how far above or below that cost you want to sit, and for how long. Almost every bad calorie target comes from collapsing those two steps into one, or from treating the first as a fact when it is a prediction with a stated error band.
The estimate itself is a short chain: a prediction equation turns your height, weight, age and sex into a basal metabolic rate, an activity factor multiplies that into a daily total, and a percentage adjustment turns the total into an intake target. Each link adds uncertainty, and the uncertainty compounds rather than cancels. By the end of the chain, a number printed to the nearest kilocalorie is honestly accurate to about plus or minus twelve per cent. This guide walks the chain, states the band at every step, sets out the floors below which a target stops being safe, and then shows how to replace the whole prediction with a measurement taken from your own body over a fortnight.
In brief
- A calorie target is built in two separate steps: estimate maintenance first, then apply a percentage adjustment to it. Combining the steps is what produces targets nobody can hold.
- The Mifflin-St Jeor equation predicts basal metabolic rate to within about ten per cent for most adults, and the activity factor adds its own error on top, so a finished target carries a band of roughly plus or minus twelve per cent.
- The 2,000 kilocalorie figure on food labels is a rounded reference value chosen so that percentages can be printed on packaging. It was never anybody's requirement and it matches very few real adults.
- Two floors bind on this site: no target below eighty per cent of estimated maintenance, and never below 1,500 kilocalories a day for men or 1,200 for women without clinical supervision.
- Fourteen days of seven-day rolling average weight will tell you more about your real maintenance than any equation, provided you correct conservatively and confirm the correction over a second fortnight.
Two questions, not one
Maintenance is what your body costs to run at your current weight, with your current activity. It is a physiological quantity. The target is what you choose to eat, which is a decision about the direction and the pace of change you want. The two get muddled constantly, usually in the form of someone deciding they will eat 1,500 kilocalories a day without ever working out what their maintenance is, and therefore without knowing whether 1,500 is a gentle ten per cent deficit or a punishing forty per cent one.
The separation matters because the two questions have different kinds of answers. Maintenance is estimated, with error, and can be measured properly with a bit of patience. The target is chosen, and the right choice depends on how much weight you have to lose, how long you intend to spend losing it, how much muscle you want to keep, and how much dietary restriction you can actually live with for months. No calculator can make that choice for you, and any calculator that pretends to has quietly made assumptions on your behalf.
The practical consequence is an order of operations. Estimate maintenance. Sanity-check the estimate against your own weight history if you have one. Choose a percentage adjustment rather than a fixed number of kilocalories, because a 500 kilocalorie cut is a sixteen per cent deficit for someone maintaining at 3,100 and a twenty-nine per cent deficit for someone maintaining at 1,700. Then check the result against the floors, which are the subject of a later section, and adjust upward if it fails them.
Estimating basal metabolic rate
Basal metabolic rate is the energy cost of being alive and at complete rest: organ function, brain activity, the maintenance of body temperature. It is the largest single component of daily expenditure for almost everyone, typically sixty to seventy per cent of the total. Measured properly it requires a person to lie still in a thermoneutral room after an overnight fast while their expired gas is analysed, which is why almost nobody has ever had theirs measured and almost everybody uses a prediction equation instead.
The equation this site uses by default is Mifflin-St Jeor, published in 1990 and fitted on healthy adults. It reads as follows, with weight in kilograms, height in centimetres and age in years:
Men: BMR = (10 x weight) + (6.25 x height) - (5 x age) + 5. Women: BMR = (10 x weight) + (6.25 x height) - (5 x age) - 161.
The reason Mifflin-St Jeor is the default rather than the older Harris-Benedict equation is a systematic review by Frankenfield and colleagues, which compared the published equations against measured resting metabolic rate and found Mifflin-St Jeor the most reliable of them in both non-obese and obese adults. Reliable, in that context, means it predicted the measured value to within ten per cent more often than its rivals did. It does not mean it predicted the measured value exactly, and the ten per cent band is the number to hold on to.
Consider a 34-year-old woman of 165 centimetres and 68 kilograms. Her predicted basal rate is (10 x 68) + (6.25 x 165) - (5 x 34) - 161, which is 680 + 1,031 - 170 - 161, or about 1,380 kilocalories a day. The ten per cent band around that runs from roughly 1,240 to 1,520. Two women identical on paper can sit at opposite ends of that band, and nothing in the equation can tell them apart, because the equation has never seen either of them. It is fitted to the average of a population, and half of any population sits on the wrong side of its own average.
| Person | Predicted BMR | Activity factor | Estimated maintenance | Honest band (plus or minus 12%) |
|---|---|---|---|---|
| Woman, 30, 160 cm, 60 kg | 1,289 kcal | 1.2 (sedentary) | About 1,550 kcal | 1,360 to 1,730 kcal |
| Woman, 45, 170 cm, 75 kg | 1,427 kcal | 1.55 (moderately active) | About 2,210 kcal | 1,950 to 2,480 kcal |
| Man, 25, 178 cm, 75 kg | 1,743 kcal | 1.375 (lightly active) | About 2,400 kcal | 2,110 to 2,690 kcal |
| Man, 60, 175 cm, 95 kg | 1,749 kcal | 1.2 (sedentary) | About 2,100 kcal | 1,850 to 2,350 kcal |
The activity factor, and where the error compounds
The activity factor is a single multiplier applied to basal rate to produce total daily energy expenditure. The conventional scale runs from 1.2 for a sedentary adult to 1.9 for someone combining physical work with serious training, and the descriptions attached to each step are vague enough that the single most common source of a wrong maintenance figure is choosing one step too high. The labels describe what you already do, not what you have resolved to start doing on Monday.
It is worth noticing that the popular 1.2 to 1.9 scale is not the only convention in use. The FAO, WHO and UNU expert consultation on human energy requirements classifies physical activity level differently, treating 1.40 to 1.69 as a sedentary or light activity lifestyle, 1.70 to 1.99 as active or moderately active, and 2.00 to 2.40 as vigorous. On that scale the familiar 1.2 sits below the floor of the lowest category. Two widely used conventions disagreeing by that much about what sedentary means is a reasonable warning against treating any multiplier as precise.
The compounding is the part people miss. If the basal rate carries about ten per cent of error and the activity factor carries its own, the product carries more than either alone. Our TDEE calculator prints a band of ten to fifteen per cent for this reason, and this guide uses twelve per cent as a middle figure when it needs to show a range. On a 2,400 kilocalorie estimate that is a spread of roughly 580 kilocalories between the top and the bottom of the band, which is more than most people's entire planned deficit.
None of that makes the estimate useless. It makes it a starting point with a stated uncertainty, which is the only honest form a starting point can take. The uncertainty is also removable, by the method in the calibration section below. What is not defensible is printing a maintenance figure to the nearest kilocalorie and then treating a forty kilocalorie difference between two calculators as though it meant anything.
Why 2,000 calories is a labelling convention
The figure of 2,000 kilocalories a day is everywhere: on nutrition labels, in the percentage daily values printed beside them, in casual conversation as though it were the human requirement. It is not a requirement. It is a reference value, chosen as a round number near the middle of adult intakes so that a packet of biscuits can print a percentage without asking who is eating them. A reference value has to be a single number, because a label has no room for a range and no way of knowing who picked up the packet.
Look back at the table above. Of the four ordinary adults in it, not one has an estimated maintenance of 2,000 kilocalories. The sedentary 30-year-old woman sits about 450 below it. The moderately active 45-year-old woman sits about 210 above. The sedentary 60-year-old man sits about 100 above, and the lightly active 25-year-old man about 400 above. Real requirements are spread across a band some 850 kilocalories wide even among four unremarkable people, and the spread widens considerably once you add very small adults, very large ones, and anyone doing physical work.
This matters practically in two directions. Someone whose true maintenance is 1,600 and who anchors on 2,000 will drift upward for years without ever feeling they overate. Someone whose true maintenance is 2,900 and who anchors on 2,000 will read a chronically low intake as normal and wonder why training feels so heavy. The label figure is doing an administrative job perfectly well. It is simply not answering the question in this guide's title, and it was never designed to.
The floors below which intake stops being safe
A deficit is a tool with a working range. Too small and it produces no detectable change against the noise of daily weight fluctuation; too large and it starts costing lean tissue, adherence, training quality and micronutrient adequacy, in roughly that order. The rails below are the ones built into our calorie deficit calculator, and whichever binds first is the one that wins.
The relative floor, at eighty per cent of estimated maintenance, exists because a percentage scales with the person. A twenty per cent deficit is a comparable stress on a 1,700 kilocalorie maintenance and a 3,100 kilocalorie one; a flat 800 kilocalorie cut is not. The absolute floors of 1,500 kilocalories for men and 1,200 for women exist because below roughly that point, meeting micronutrient reference intakes from ordinary food becomes genuinely difficult rather than merely inconvenient, and because those are the figures used in practice to mark where a diet stops being self-managed and starts needing supervision.
Very low energy diets below those floors are real, they work, and they have a legitimate clinical place. Their place is under a doctor or dietitian who can monitor electrolytes, medication interactions and micronutrient status, and who has actually met the patient. That is a different thing from a web form arriving at 900 kilocalories a day because the arithmetic allowed it.
One further floor is worth stating even though it is not a calorie number. A goal weight below a body mass index of 18.5 for your height is not a target this site will produce, because the healthy weight range has a bottom as well as a top and the bottom is where most of the harm in this field happens. If a target requires you to go below it, the target is wrong rather than ambitious.
| Rail | Value | What it protects | What happens without it |
|---|---|---|---|
| Relative floor | No lower than 80% of estimated maintenance | Lean tissue, training quality and adherence, scaled to the individual | A fixed cut that is mild for a large person becomes severe for a small one |
| Absolute floor | 1,500 kcal a day for men, 1,200 for women | Micronutrient adequacy from ordinary food | Intakes that require supplementation and monitoring to remain nutritionally complete |
| Weekly rate cap | No more than 1% of body weight a week | The proportion of the loss that is fat rather than lean mass | Faster scale movement bought with a larger share of lean tissue |
| Goal weight floor | Not below a BMI of 18.5 for your height | The bottom of the healthy weight range | A tool that will happily compute a target below the range it is meant to be screening for |
Correcting the estimate from a fortnight of weight data
Everything above is prediction. Your own body weight over time is measurement, and it beats every equation in this guide, because it is the only instrument that has actually met you. The procedure is simple and the discipline it requires is in the tracking rather than the arithmetic.
Eat at the estimated maintenance figure for fourteen days without changing your training, weighing food rather than judging it by eye, and logging everything including cooking oils and drinks. Weigh yourself every morning under identical conditions, after the bathroom and before food or drink, and record only the seven-day rolling average. Compare the average on day seven with the average on day fourteen. The direction and size of that change is the correction.
Converting the change into kilocalories uses the conventional figure of about 7,700 kilocalories per kilogram of mixed tissue. A loss of half a kilogram over the fortnight is a quarter of a kilogram a week, which is roughly 1,925 kilocalories a week, or 275 kilocalories a day of unintended deficit, meaning your true maintenance is about 275 above the estimate. A gain of the same size means the reverse.
There is an important honesty caveat on a fortnight, and it is the reason this section says correct conservatively. Over short windows a large share of any weight change is not fat. Research measuring the composition of two-week weight changes in free-living adults found the great majority of it was fat-free mass, at an energy density far below 7,700 kilocalories per kilogram. Apply the 7,700 conversion to a two-week window and you will systematically overstate the correction. The defensible approach is to apply half of the implied correction, then run a second fortnight at the new figure and see whether the average has flattened. Two conservative corrections landing in the same direction are worth far more than one confident one.
Recalibrate on a schedule rather than on frustration: after every five kilograms of weight change, or any substantial change in job, training or routine. Between those points, resist adjusting from single weeks. A seven-day average has to move consistently for three or four weeks before it is telling you about energy balance rather than about water, sodium, glycogen or the contents of your gut.
- Set the trial intake at the calculated maintenance figure and do not change training, sleep or step count for the fortnight.
- Weigh food on scales. Estimated portions are the single largest source of error in this entire procedure, and they err in one direction.
- Weigh yourself daily under identical conditions and record only the seven-day rolling average. Never react to one morning.
- On day fourteen, compare the day-seven average with the day-fourteen average, in kilograms, to one decimal place.
- Convert at roughly 7,700 kcal per kilogram to get the implied daily error, then apply half of it.
- Hold the corrected figure for a second fortnight and confirm the average is flat before treating it as your maintenance.
- Write the raw daily weights down. If a fortnight looks strange later, only the raw numbers can tell you why.
| Change across the fortnight | Implied weekly rate | Implied daily error at 7,700 kcal/kg | What to do |
|---|---|---|---|
| No change, within 0.3 kg | Flat | None detected | Treat the estimate as usable and start the deficit from it |
| Down 0.5 kg | 0.25 kg a week | About 275 kcal a day below maintenance | Raise the maintenance figure by about 140 kcal and re-run |
| Down 1.0 kg | 0.5 kg a week | About 550 kcal a day below maintenance | Raise by about 275 kcal and re-run; check the intake log for missed days |
| Up 0.5 kg | 0.25 kg a week | About 275 kcal a day above maintenance | Lower the maintenance figure by about 140 kcal and re-run |
| Up 1.0 kg | 0.5 kg a week | About 550 kcal a day above maintenance | Lower by about 275 kcal, and check for a recent change in salt, carbohydrate or training |
Turning maintenance into a target
With a calibrated maintenance figure, the second question becomes easy to answer well. For weight loss, ten to twenty per cent below maintenance is the range that reliably produces change while preserving lean tissue and remaining liveable. For lean gain, ten per cent above is usually enough, and the weekly rate matters more than the surplus: a rate capped near half a per cent of body weight a week keeps most of the gain as tissue you wanted.
Return to the 34-year-old woman from earlier, with a predicted basal rate of 1,380 and a lightly active factor of 1.375, giving an estimated maintenance of about 1,900 kilocalories. A twenty per cent deficit puts her target at 1,520, which clears the 1,200 absolute floor comfortably and sits exactly on the eighty per cent relative floor. The implied deficit of 380 kilocalories a day is about 2,660 a week, or roughly 0.35 kilograms a week at the conventional conversion, which is well inside a one per cent of body weight cap of 0.68 kilograms. All four rails pass, which is the check worth running before starting rather than three weeks in.
Two habits separate targets that work from targets that do not. The first is holding the number long enough for it to say something: three to four weeks of consistent seven-day averages, not one frustrating Tuesday. The second is adjusting the target rather than abandoning it. When progress stalls genuinely, the fix is usually a hundred kilocalories, a recovered step count or an honest look at the weekend logging, and almost never a dramatic new approach.
It is also worth deciding in advance what happens at the end. A deficit is a temporary state with a defined exit, and the exit is a deliberate return to maintenance at the new, lower body weight, which will be a smaller number than the maintenance you started from. Planning that return is the part most people skip, and it is the part that decides whether the loss stays lost.
Frequently asked questions
Is 1,200 calories a day enough?
For most adult women it is the recognised lower limit rather than a sensible target, and for men the equivalent figure is 1,500. Below those levels, meeting reference intakes for iron, calcium, and the B vitamins from ordinary food becomes genuinely hard, and lean tissue loss climbs as a share of total weight lost. They are also the thresholds at which a diet stops being reasonably self-managed and starts warranting supervision from a clinician who can monitor electrolytes and medication. If a calculator produces a target below them, the correct response is to raise the target and accept slower progress rather than to accept the number.
Why do different calculators give me different calorie numbers?
Because they use different prediction equations, different activity factor scales, and sometimes different rounding, and because none of those choices has a single correct answer. Mifflin-St Jeor, Harris-Benedict, Katch-McArdle and Owen all predict basal rate from slightly different data and disagree by a hundred kilocalories or more on the same person. The activity multiplier then amplifies that disagreement. A spread of two or three hundred kilocalories between reputable calculators is normal and expected, which is the clearest possible argument for treating any of them as a starting estimate and calibrating against your own weight data.
Should I eat back the calories my watch says I burned?
Usually not, and the reason is double counting rather than distrust of the watch. If your maintenance figure came from a basal rate multiplied by an activity factor, your training is already inside that figure, and adding session calories on top counts it twice. There is a second problem: wearables and gym machines report gross energy cost, which includes the energy you would have spent resting anyway, so the genuine additional cost of a session is meaningfully smaller than the number displayed. A single weekly maintenance figure, calibrated against your own weight trend, avoids both problems.
How accurate is a calculated calorie target?
The basal rate underneath it is accurate to roughly ten per cent for most adults, and the activity factor adds further uncertainty, so a finished target is honestly accurate to about plus or minus twelve per cent. On a 2,400 kilocalorie figure that is a spread of nearly 600 kilocalories from the bottom of the band to the top, which is larger than most planned deficits. That is not a reason to ignore the calculation; it is a reason to treat it as a starting point, hold it for a fortnight, and correct it from your own seven-day weight averages rather than from a better equation.
Why has my weight not moved even though I am in a deficit?
In the first three weeks, the overwhelmingly likely explanations are water and measurement rather than metabolism. Glycogen shifts, sodium changes, the menstrual cycle, a new training stimulus and simple gut contents all move scale weight by amounts that dwarf a fortnight of fat loss. The second most likely explanation is that intake is higher than recorded, because under-reporting in free-living adults is large and systematic, and cooking oils, drinks and unweighed portions are the usual gaps. Genuine metabolic adaptation exists but is a slow, modest headwind, not something that appears in week three.
Put it into practice
Run your own numbers through the TDEE calculator, the BMR calculator, the calorie deficit calculator and the macros calculator. Related reading: TDEE explained, Setting your macros and Adaptation and plateaus.
Sources
- Mifflin MD, St Jeor ST, Hill LA, Scott BJ, Daugherty SA, Koh YO. A new predictive equation for resting energy expenditure in healthy individuals. Am J Clin Nutr 1990;51:241-7. doi.org/10.1093/ajcn/51.2.241
- Frankenfield D, Roth-Yousey L, Compher C. Comparison of predictive equations for resting metabolic rate in healthy nonobese and obese adults: a systematic review. J Am Diet Assoc 2005;105:775-89. doi.org/10.1016/j.jada.2005.02.005
- FAO/WHO/UNU. Human energy requirements: report of a joint FAO/WHO/UNU expert consultation. FAO Food and Nutrition Technical Report Series 1. Rome, 2004. www.fao.org/4/y5686e/y5686e07.htm
- Hall KD, Sacks G, Chandramohan D, et al. Quantification of the effect of energy imbalance on bodyweight. Lancet 2011;378:826-37. doi.org/10.1016/S0140-6736(11)60812-X
- Bhutani S, Kahn E, Tasali E, Schoeller DA. Composition of two-week change in body weight under unrestricted free-living conditions. Physiol Rep 2017;5:e13336. doi.org/10.14814/phy2.13336
- Institute of Medicine. Dietary Reference Intakes for Energy, Carbohydrate, Fiber, Fat, Fatty Acids, Cholesterol, Protein, and Amino Acids. Washington DC: National Academies Press, 2005. doi.org/10.17226/10490
Cite this page
Quoting a figure from here in an article, a report or a piece of coursework? Use whichever of these your style guide asks for.
- APA
- Campbell, R. (2026). How many calories should I eat?. Body Stats. https://bodystats.co/app/guides/how-many-calories-should-i-eat
- Plain text
- “How many calories should I eat?”, Body Stats, last updated 13 September 2026, https://bodystats.co/app/guides/how-many-calories-should-i-eat
This guide is informational and educational, not medical advice. Formula details live on the methodology page; see also the medical disclaimer.
Last updated . Written by Rick Campbell; not medically reviewed. See review status.