Keto macros: how to calculate yours, and the protein mistake most people make

To calculate keto macros, work in grams and in this order: estimate your daily calories (resting metabolic rate times an activity factor, minus a deficit if you want to lose weight), set protein at roughly 1.2 to 1.6 g per kg of lean or goal body weight, cap net carbs at about 20 to 30 g, and let fat fill whatever calories are left. The common mistake is doing it backwards: picking a "75/20/5" ratio first and letting protein shrink every time calories drop.

Below is the full method, the reasoning behind each number, one worked example you can copy, and the mistakes that make keto macros quietly fail.

Why "75/20/5" is the wrong place to start

Most keto calculators open with a ratio, typically 70 to 75% fat, 20 to 25% protein and 5 to 10% carbs. The ratio is a fine description of what a keto day often ends up looking like. It is a bad way to set one.

Here is the problem. Percentages scale with calories, and your body's protein needs do not. If you cut to 1,400 calories and keep protein at 20%, you get 280 calories of protein, which is 70 g. For someone with a goal weight of 70 kg, that is 1.0 g/kg, while you are in a deficit, which is exactly when your body is most willing to give up muscle. Fat stays at 70%-plus of a smaller number, so the plate still "looks" keto. It just is not protecting you.

So: grams first, percentages last.

Step 1: estimate your resting metabolic rate (Mifflin-St Jeor)

Your basal or resting metabolic rate (BMR or RMR) is roughly what you burn lying still. The most widely used formula is Mifflin-St Jeor, published in 1990 from indirect calorimetry measurements in 498 healthy adults aged 19 to 78, both normal-weight and obese (Mifflin et al., AJCN):

• Men: 10 × weight (kg) + 6.25 × height (cm) − 5 × age (years) + 5
• Women: 10 × weight (kg) + 6.25 × height (cm) − 5 × age (years) − 161

Why this one? A systematic review comparing the four most common equations (Harris-Benedict, Mifflin-St Jeor, Owen and WHO/FAO/UNU) found Mifflin-St Jeor the most reliable, predicting RMR within 10% of the measured value in more people than any other, with the narrowest error range (Frankenfield et al., J Am Diet Assoc). The same review is blunt about the limits: there are "noteworthy errors" for individuals, and older adults and US ethnic minorities were underrepresented in the validation data.

Translation: it is a good first guess, not a measurement.

Watch out: The original 1990 paper found the older Harris-Benedict equations overestimated measured resting energy expenditure by about 5%. If you got your numbers from an old calculator, your starting calories may be a little high.

Step 2: multiply by an activity factor

Your total daily energy expenditure (TDEE) is BMR times an activity multiplier. These are the conventional ones most calculators use. They are rules of thumb, not lab values:

The negative to know about: people overestimate their activity far more often than they underestimate it. Three gym sessions and a desk job is "lightly active" for most people, not "moderately active". If you are torn between two levels, pick the lower one.

Step 3: subtract a sensible deficit (if you are losing weight)

The National Heart, Lung, and Blood Institute's guidance is that cutting 500 to 1,000 calories a day from your intake supports a loss of about 1 to 2 pounds a week (NHLBI). For smaller or older people, the lower end is usually more realistic: a 1,000-calorie cut from a 1,800-calorie TDEE is a big, hard-to-sustain gap.

Keto does not exempt you from energy balance. It can make a deficit easier to live with, because many people find high-fat, high-protein meals filling. But if your calories are above your TDEE, you will not lose fat no matter how low your carbs are. (More on that in Not losing weight on keto? 11 hidden reasons.)

Step 4: set protein in grams, from body weight

This is the step that matters most, and the one most calculators get wrong.

The floor. The Recommended Dietary Allowance is 0.8 g of protein per kg of body weight. Harvard Health describes it as roughly the minimum you need to keep from getting sick, not the amount you should aim for (Harvard Health).

What the evidence suggests above the floor:

• A meta-analysis of 49 trials and 1,863 participants found that, during resistance training, gains in fat-free mass stopped increasing beyond a total protein intake of about 1.6 g/kg/day (Morton et al., BJSM).
• In adults over 50 losing weight, a meta-analysis of 20 randomized trials found that higher-protein diets (at least 25% of energy or 1.0 g/kg/day) retained more lean mass and lost more fat (Kim et al., Nutrition Reviews).
• The International Society of Sports Nutrition recommends 1.4 to 2.0 g/kg/day for exercising people building or keeping muscle (ISSN position stand).

A practical keto range for most adults is therefore about 1.2 to 1.6 g/kg, toward the top if you lift or are in a meaningful deficit.

Which kilograms? If you have a lot of body fat, multiplying by your current total weight overshoots, because fat tissue does not need protein the way muscle does. Two better options:

• Goal weight: a realistic target weight, not a fantasy one.
• Lean body mass: weight × (1 − body-fat fraction). With lean mass, use the higher end, about 1.6 to 2.0 g/kg, since you have already removed the fat.

Bunny tip: Set protein as a daily gram number and hit it first, split across two or three meals. Then fill in fat. If you build the day around fat, protein is the macro that gets squeezed out.

Step 5: cap net carbs

Net carbs are total carbohydrate minus fiber. Ketogenic diets are generally defined by very low carbohydrate intake, typically about 20 to 50 g a day (StatPearls, NIH). For most people starting out, a strict cap of 20 to 30 g net carbs is the simplest way to reliably reach ketosis. Some people can go higher and stay in ketosis; you only find that out by testing (see How to know you're in ketosis).

Two negatives to plan for:

• Hidden carbs in sauces, "keto" snacks and vegetables add up fast. A 25 g cap leaves little room for guesswork.
• Sugar alcohols are not all equal. Subtracting all of them from total carbs flatters the label, and maltitol is the usual offender (see Sugar alcohols on keto).

Step 6: fat is the remainder

Fat is whatever calories are left after protein and carbs, divided by 9 (fat has 9 calories per gram; protein and carbohydrate have about 4).

Treat this number as a ceiling, not a quota. If you are losing weight and you are not hungry, you do not need to add butter to "hit" fat. That is the single most common way people turn a deficit into maintenance on keto. If you are hungry, fat is the flexible lever to pull.

A worked example

Meet a 40-year-old woman, 168 cm tall, weighing 85 kg, lightly active, with a goal weight of 68 kg.

BMR (Mifflin-St Jeor, female): 10 × 85 + 6.25 × 168 − 5 × 40 − 161 = 850 + 1,050 − 200 − 161 = 1,539 calories

TDEE: 1,539 × 1.375 = about 2,116 calories

Target with a 500-calorie deficit: 2,116 − 500 = about 1,616 calories

Protein: 1.6 g × 68 kg goal weight = 108.8, round to 110 g → 110 × 4 = 440 calories

(Check against lean mass: at about 35% body fat, lean mass is roughly 85 × 0.65 = 55 kg, and 55 × 2.0 = 110 g. Same answer.)

Net carbs: cap at 25 g → 25 × 4 = 100 calories

Fat: 1,616 − 440 − 100 = 1,076 calories ÷ 9 = about 120 g (as an upper limit)

Still unmistakably keto, but protein is 27%, not 20%. A 75/20/5 template at the same calories would have given her about 81 g of protein, roughly 1.2 g/kg of goal weight: a worse deal in a deficit.

The protein mistake, explained: "too much protein kicks you out of ketosis"

The fear goes like this: excess protein turns into glucose through gluconeogenesis, so eating more protein will knock you out of ketosis. That is why many keto guides tell you to keep protein "moderate", and why people under-eat it.

What the evidence actually shows is more reassuring:

• Gluconeogenesis is largely demand-driven. Your body makes glucose for the tissues that need it, whether or not you eat a lot of protein. In a tightly controlled crossover study, healthy men on a high-protein, carbohydrate-free diet (30% protein, 70% fat) did produce a larger share of their glucose via gluconeogenesis, but their total glucose production was lower than on a normal high-carb diet (181 vs 226 g a day) (Veldhorst et al., AJCN). Protein did not flood them with sugar.
• Research keto diets are not low-protein. A 2024 review of well-formulated ketogenic diets notes that very low carbohydrate studies typically report protein intakes of 0.6 to 1.4 g/kg a day, similar to the standard American diet, while maintaining nutritional ketosis, and below the roughly 2.0 g/kg threshold usually considered "high protein" (Athinarayanan et al., BMJ Open Diabetes Research & Care).

So the real-world mistake is not eating too much protein. It is eating too little, because the macro template said 20% and the fear said "careful".

Watch out: "Protein is fine" has an important exception. If you have chronic kidney disease, protein targets are a medical decision. Do not raise your protein, or start keto at all, without your nephrologist or doctor. The same applies if you take insulin or sulfonylureas (keto can lower blood sugar quickly and doses may need changing), are pregnant or breastfeeding, or have a history of eating disorders, where tracking numbers can do more harm than good.

Other macro mistakes that stall people

1. Recalculating never. Your BMR drops as you lose weight. Recalculate after every 5 kg or so lost.
2. Trusting the formula over the scale. If your weight trend is flat for three to four weeks while you are logging honestly, your real TDEE is lower than the estimate. Drop about 100 to 200 calories, usually from fat.
3. Counting "net" carbs on products that subtract maltitol or other high-impact sweeteners. The label math is legal but optimistic.
4. Treating fat as a target. Fat bombs and bulletproof coffee are calorie-dense. They are a tool for hunger, not homework.
5. Ignoring liquid and "invisible" calories. Cream, oils and cooking butter rarely get logged accurately.

When to adjust your macros

Give the numbers two to four weeks with consistent logging before you judge them. Then:

• Losing steadily and feeling fine: change nothing.
• Not losing, and logging is accurate: trim fat by 10 to 20 g a day.
• Constantly hungry or weak in the gym: check that protein is actually being hit; then add some fat.
• Not in ketosis by a blood test: look for hidden carbs before touching protein.

Track it with KetoBunny: Once you have your numbers, the hard part is hitting them on real food. KetoBunny lets you photo-log a meal, shows net carbs, protein and fat against your targets, and gives an estimate of where you probably are with ketosis. It is an estimate, not a blood test, but it makes it obvious on which days protein slipped or carbs crept up.

The short version

Calories first, then protein in grams, then a net-carb cap, then fat as the flexible remainder. Mifflin-St Jeor is the best common starting estimate, and it is still only an estimate. Protein is the macro most likely to be set too low on keto, not too high. And if you have kidney disease, take diabetes medication, are pregnant, or have an eating-disorder history, set your macros with a doctor rather than a calculator.