Maintenance calories without an app

01What are maintenance calories?

Maintenance is the intake at which your weight holds steady — the reference line every cut, bulk, or recomp is measured against. Apps sell you this number with decimal-point confidence. You can build a better one with a pencil and a bathroom scale, and you'll know exactly how much to trust each digit.

02How do you estimate maintenance with Mifflin-St Jeor?

Start with resting metabolic rate — what you'd burn doing nothing all day. The Mifflin-St Jeor equation, built from indirect-calorimetry measurements of 498 adults, is the pencil version [1]: bodyweight in kg times 10, plus height in cm times 6.25, minus age in years times 5, then plus 5 for men or minus 161 for women [1].

A systematic review put the four most-used formulas head-to-head; Mifflin-St Jeor was the most reliable — the most likely to land within 10% of a measured value, with the narrowest error range [2]. The century-old Harris-Benedict runs about 5% hot against measurement [1]. But read the best case honestly: 10% of that 1,733 is ±170 kcal, before you've guessed at activity. An equation gets you the neighborhood, not the address.

03Why do activity multipliers fail?

Resting burn isn't maintenance — every calculator multiplies it by an activity factor, and this is where estimates go to die, because the factor is a self-graded adjective. The government's own calorie table makes the stakes plain: for a 40-year-old man, sedentary reads 2,400 kcal, moderately active 2,600, active 2,800 [3]. One honest disagreement about your own week moves the target 200–400 kcal.

And the adjectives are stricter than they feel: moderately active already means the equivalent of walking 1.5–3 miles a day on top of normal life [3]. Most desk weeks are sedentary by the table's own footnote. So pick the lower adjective, call the result a draft, and let the scale vote.

04How does the two-week scale method work?

Weigh in every morning for two weeks — same scale, same conditions — while eating the way you actually eat and writing down roughly what goes in. The office cake, the kids' leftovers, all of it: that's the life the number has to fit. Average week one. Average week two. If the averages match, your logged intake is your maintenance — measured on your body, your commute, your kitchen. If the trend moved, adjust intake 100–200 kcal against the drift and run it again.

This beats every formula for one reason: it isn't a prediction. It's a measurement of the only subject who matters. One warning from the modeling literature — bodyweight answers intake changes slowly, with half-times around a year, and the old 3,500-kcal-per-pound rule ignores that adaptation and has led to drastically overestimated expectations [4]. Use the trend to steer your next 100–200 kcal step, not to forecast a February version of yourself.

05Can you trust your watch's calorie burn?

No — and this is measured, not opinion. In a Stanford validation of seven wrist-worn devices against indirect calorimetry, none estimated energy expenditure within 20%; median errors ran from 27% on the best device to 93% on the worst [5]. A systematic review spanning 158 studies reached the blunt version: for energy expenditure, no brand was accurate [6].

Steps and heart rate hold up far better in the same review [6] — wear the watch for those. But a burn number that can miss by a quarter is not an input to arithmetic. If your maintenance came from the two-week trend, your training is already priced in. Nothing to eat back, nothing to sync.