Calories: how to estimate your maintenance

Why every calculator is wrong

Maintenance calories — the intake at which your weight holds steady — depend on your size, body composition, activity, non-exercise movement, genetics, and how efficiently you absorb food. Equations capture some of that and estimate the rest.

The result is that a good equation gets most people within about 10%, which sounds fine until you translate it: for a 2,600 kcal maintenance, 10% is 260 kcal — larger than the deficit many people are trying to run. Two people of identical weight, height, age and stated activity can differ by 400 kcal or more, largely through spontaneous daily movement.

Getting a starting number

The simplest approach that works is body weight in kilograms times a multiplier:

If you prefer more structure, calculate resting metabolic rate with the Mifflin-St Jeor equation and multiply by an activity factor:

• Men: (10 × kg) + (6.25 × cm) − (5 × age) + 5
• Women: (10 × kg) + (6.25 × cm) − (5 × age) − 161
• Then multiply by 1.4 (sedentary plus some training) to 1.8 (very active).

Both methods land in the same territory. Neither is more accurate than the other in a way that matters, because the error comes from the activity multiplier rather than the equation.

The only accurate method

You do not calculate maintenance. You measure it.

1. Eat roughly your estimate for two full weeks, tracking honestly. Not approximately — actually weighing food, at least for these two weeks.
2. Weigh yourself daily, same conditions: morning, after the toilet, before eating.
3. Take the average of week one and the average of week two. Never compare single days — see why the scale lies day to day.
4. Compare the two weekly averages. Stable within about 0.3 kg means you have found maintenance. Rising means your maintenance is higher than you are eating. Falling means lower.
5. Adjust by roughly 100–200 kcal and repeat if needed.

Two weeks of this gives you a number that no calculator can. It is also the single most useful piece of nutritional information you will ever have about yourself, and it stays roughly valid for as long as your weight and activity do.

Why your maintenance is a moving target

Maintenance is not a fixed constant, which is why people who found their number two years ago are often confused when it stops working.

• It falls as you lose weight. A smaller body costs less to run. Roughly 20–25 kcal per kg lost, before any adaptive effects.
• Non-exercise movement adapts. In a deficit, people unconsciously fidget less, walk less and take the lift. This is the largest and least visible component of change.
• It rises as you gain muscle, though modestly — muscle costs around 13 kcal per kg per day at rest, so 5 kg of new muscle is about 65 kcal, not the hundreds sometimes claimed.
• Activity changes seasonally. Most people move considerably more in summer than in winter, without noticing.
• Adaptive thermogenesis is real but small. After roughly 14 kg of weight loss, measured adaptation was about 92 kcal/day below prediction — falling to about 38 kcal/day after four weeks of weight stabilisation. Meaningful, not metabolism-breaking.

Recheck your number after any substantial weight change, or any lasting change in job or activity.

Setting intake from maintenance

Once you have a real number, the phases follow from it:

• Fat loss: maintenance minus 300–500 kcal, targeting 0.5–1.0% of body weight lost per week — see building a calorie deficit you can sustain.
• Muscle gain: maintenance plus 200–350 kcal, targeting 0.25–0.5% of body weight gained per week — see eating for muscle gain without excess fat.
• Maintenance: the number itself, which is a legitimate and underused phase — see maintenance: the phase nobody plans.

Note the asymmetry: fat loss tolerates a larger adjustment than muscle gain does, because excess surplus becomes fat quickly whereas a moderate deficit is well tolerated.

Common questions

My tracker says I burn 3,200 calories. Is that right?

Almost certainly not. Wearable energy expenditure estimates are frequently off by 20% or more, usually high. Use your weight trend instead.

Why is my maintenance lower than the calculator says?

Usually under-reported eating, over-estimated activity, or both. Occasionally a genuinely low activity level. Two weeks of careful tracking distinguishes these.

Should I eat more on training days?

You can, and some people prefer it. Weekly total is what determines the trend, so it is a preference rather than a requirement.

Do I need to count on maintenance too?

Not forever. Counting for a few weeks builds an accurate sense of portions that persists long after you stop — see tracking calories without becoming obsessive.

How much does metabolism differ between people?

Resting rate varies by roughly 5–10% between people of the same size. Total expenditure varies far more, mostly through daily movement.

Is a "damaged metabolism" real?

Not in the way it is usually meant. Adaptation exists and is modest; the larger effect is reduced spontaneous movement plus under-reported intake — see reverse dieting after a cut.

The practical version

• Start at body weight (kg) × 29–33 for a desk job plus three or four sessions a week.
• Every calculator is an estimate; the error is usually in the activity multiplier.
• Measure your real maintenance: two weeks of honest tracking plus daily weigh-ins.
• Compare weekly averages, never single days.
• Maintenance falls as you lose weight, mostly via reduced spontaneous movement.
• Adaptive thermogenesis is real but modest — roughly 38 kcal/day once weight stabilises.