Worked example: 30-year-old man, 80 kg, 180 cm, 18% body fat, moderately active
| Formula | Basal (BMR) | At 1.55 activity |
|---|---|---|
| Mifflin-St Jeor (1990) | 1,780 kcal | 2,759 kcal |
| Harris-Benedict, Roza-Shizgal revision (1984) | 1,854 kcal | 2,874 kcal |
| Katch-McArdle (1996) | 1,787 kcal | 2,770 kcal |
The three formulas span 74 kcal at basal, about 4%. That is a smaller disagreement than you might expect, and it points at something more important: the formula is not where your uncertainty lives. The activity multiplier is.
The activity multiplier is the real variable
Moving one step on the activity scale — from lightly active (1.375) to moderately active (1.55) — changes this person's maintenance from 2,448 to 2,759 kcal. That is a 311 kcal swing from a dropdown you filled in by guessing. It dwarfs the 74 kcal disagreement between the three published formulas that everyone argues about.
The practical consequence: do not agonise over which formula to pick. Pick Mifflin-St Jeor, be conservative with the activity level, and then correct from reality after two weeks of scale weight. Nobody's true maintenance was ever discovered by a calculator; it is discovered by eating a set amount for a fortnight and watching what happens.
Which formula, if you want the short answer
Mifflin-St Jeor unless you have an actual measured body-fat percentage, in which case Katch-McArdle. Katch-McArdle is the only one of the three that knows the difference between 80 kg of mostly muscle and 80 kg of mostly not, which makes it the better formula for lean and for very heavy people alike. But if you are guessing your body fat, Katch-McArdle is worse than Mifflin-St Jeor, because you have swapped a measured variable for an invented one. Harris-Benedict is included because it is the oldest and most cited; it runs consistently high on modern populations, which is why Mifflin-St Jeor replaced it.
Methodology
Mifflin-St Jeor: BMR = 10w + 6.25h − 5a + 5 (men) or −161 (women), with weight in kg, height in cm, age in years. Harris-Benedict in the Roza-Shizgal 1984 revision. Katch-McArdle: BMR = 370 + 21.6 × lean body mass in kg. Maintenance is BMR multiplied by an activity factor from 1.2 (sedentary) to 1.9 (very heavy training or physical work).
Limitations, stated plainly
All three formulas predict a population average, and individual metabolic rate scatters around it by roughly ±10%. For the example above, that means true maintenance could genuinely be 2,480 or 3,040 kcal. On top of that sits the activity multiplier, which is self-reported and systematically over-estimated by almost everyone. The formulas are also least accurate at the extremes — very lean, very heavy, very old, very young — because that is where the source populations were thinnest.
Use the number as a starting point, then adjust using two weeks of actual scale weight measured under the same conditions.
The floor that applies to everything downstream of this number
Maintenance is the figure every other calorie target on this site is built from, so the limit belongs here too. No target derived from this page — a cut, a macro split, anything — should send you below roughly 1,500 kcal a day if you are a man or roughly 1,200 if you are a woman. Below those figures it becomes difficult to meet micronutrient requirements from food at all, and the answer stops being one a general-purpose calculator can responsibly give. If the arithmetic takes you there, take a smaller deficit and a longer timeline, or speak to a registered dietitian or your doctor, who can look at your actual situation rather than at four inputs.
This is a fitness estimate, not medical advice, and not a substitute for a professional who can actually see you. If you have any condition affected by what you eat, or any history of disordered eating, that conversation belongs with a doctor or a dietitian rather than with a web page.