Macro Split: How Much Protein, Carbs and Fat

Stacked bar of a 2,400 calorie day split into 170 g of protein, 77 g of fat and 257 g of carbohydrate

Your protein target comes from your body weight. 1.6 to 2.2 g per kilogram a day, whatever your calorie budget happens to be that month. Percentage splits like 40/30/30 tie protein to the calorie budget instead, and those two numbers move independently of each other.

30% protein at 3,000 calories is 225 g. The same 30% at 1,400 calories is 105 g. At 85 kg the first figure is generous and the second leaves your muscle exposed in a deficit. The percentage held still and so did your requirement, which is the problem with setting protein as a share of calories.

Set protein first, in grams per kilogram

Research settles on 1.6 to 2.2 g per kg of body weight per day for anyone training. Go toward the top of that when you are in a deficit or already lean, because that is when your muscle is most at risk.

At 85 kg, that is 136 to 187 g a day. At 4 calories per gram, 544 to 748 calories. How much protein you need a day covers how to hit it and what happens when you miss.

Bar chart showing 30 percent protein giving 225 g at 3,000 calories and 105 g at 1,400 calories, against an actual need of 136 to 187 g
Your protein need tracks your body weight. Your calorie budget tracks your goal. Tying one to the other breaks at both ends.

Set fat second, also per kilogram

Fat has a floor, not a target. Below roughly 0.5 g per kg, or under about 20% of total calories, most people run into hormonal problems and start losing their grip on the fat-soluble vitamins.

0.8 to 1.0 g per kg is a comfortable working range. At 85 kg that is 68 to 85 g, which at 9 calories per gram is 612 to 765 calories.

Carbohydrate gets the rest

Your body makes glucose from other sources, so carbohydrate has no minimum requirement the way protein and fat do. What it has is a performance role. It fuels hard training and it refills muscle glycogen.

Carbohydrate takes whatever is left. Total calories, minus protein calories, minus fat calories, divided by four.

Worked example

An 85 kg man maintaining on 2,900 calories, eating 2,400 to lose fat:

  • Protein at 2.0 g/kg: 170 g, 680 calories
  • Fat at 0.9 g/kg: 77 g, 693 calories
  • Carbohydrate: 2,400 minus 1,373, which is 1,027 calories, so 257 g

That works out to 28% protein, 29% fat, 43% carbohydrate. You could have started from those percentages, and you would have been guessing at them. This way the numbers came out of his body weight.

Get your total first. The calorie calculator gives you maintenance and a target from your own measurements, and the case against a flat 2,000 calories a day explains why the label figure was never yours.

When to change the split

Very low calorie targets. At 1,400 calories, protein at 2.0 g/kg for an 85 kg person already eats half the budget. Hold the protein and cut the carbohydrate, because protein is what protects the muscle you are trying to keep.

Endurance training. More than an hour of hard aerobic work most days pushes your carbohydrate requirement up. Take it from fat and not from protein.

Gaining muscle. In a surplus, send the extra calories mostly to carbohydrate. Protein above 2.2 g/kg has shown no extra benefit in trials. How big a calorie surplus should be has the figures.

How much precision is worth it

Hit your protein within about 10 g. The split between carbohydrate and fat matters much less than people think, and several trials comparing low fat with low carbohydrate diets at matched calories and protein have landed in the same place.

Pick the split you will follow. Someone who feels better on more fat and someone who trains better on more carbohydrate will both do fine. The one who quits in three weeks will not.

Whichever split you pick, you are probably missing it. People underreport their intake by 20 to 40%, and the error lands mostly on fats and oils. How wrong calorie tracking usually is covers where it comes from.

Scope and correction note, 9 November 2026: the ranges quoted here come from the American Council on Exercise and from published research, and they describe averages, not any individual. Nothing on this page is medical advice. If a figure here is wrong, tell us and we will correct it and date the correction. See the editorial policy and medical disclaimer.

Run the numbers

Put this to work: