How to Set Your Fat Intake Without Guessing

Protein gets a number. Carbs get a number, usually tied to how much you train. Fat gets whatever calories are left over. That works more often than it should, but it fails badly in one direction, and the people it fails are usually the ones dieting hardest.

So: how much fat should you actually eat, and how low is too low?

Fat is the macro nobody sets on purpose

Almost every macro split is built the same way. Set protein from bodyweight. Set calories from an estimate of maintenance. Fill the remainder with carbs and fat in whatever ratio you like the taste of. If you have read how the three macros fit together, you already know the arithmetic works out regardless of how you split that remainder, because calories are calories.

The problem is that the remainder shrinks. Hold protein fixed, cut calories by 500, and something has to give. Carbs usually get defended because people can feel a low-carb training session. Fat does not announce itself the same way, so it quietly gets squeezed toward zero over the course of a long diet. Nobody decides to eat 25 grams of fat a day. They arrive there.

The published guidance is thinner than you would expect

Protein has clear numbers you can look up. Fat does not, and this is not you failing to find them.

A 2025 scoping review by Delany and colleagues in Sports Medicine went through 73 consensus statements, position stands and practice guidelines from international expert groups, looking specifically at what they say about changing body mass and composition. On fat, only 13 of those documents said anything at all. Between them they offered seven different targets, spanning 15 to 30 percent of energy intake, and three warned against excessive restriction below 15 percent. Their own summary is that the guidance pays minimal attention to fat and that the language used is often vague.

That is worth sitting with. The reason you cannot find a crisp answer is that the expert groups have not agreed on one either. Across the documents, lower-end recommendations generally landed around 15 to 20 percent, but the targets varied and no single clinical threshold emerged.

The testosterone question, honestly

This is the part most people are really asking about, so here is the state of it.

In 2021, Whittaker and Wu pooled six intervention studies covering 206 men and found significant decreases in total testosterone, free testosterone, urinary testosterone and dihydrotestosterone on low-fat diets compared with high-fat diets. The effect on total testosterone was a standardized mean difference of -0.38. The authors were careful in their own conclusion: low-fat diets appear to decrease testosterone in men, but further randomized controlled trials are needed to confirm it.

Those trials have accumulated. In 2025, Soltani and colleagues ran a larger meta-analysis of 11 randomized controlled trials with 888 participants and found no significant difference in testosterone, estradiol, estrone, SHBG, DHEA, progesterone or androstenedione between low-fat and high-fat diets. They also tempered that: small number of studies, low certainty of evidence.

Two meta-analyses, two answers. What I take from the pair is that if an effect exists it is small, it is not reliably detectable, and it is not the strongest argument for keeping fat above a floor. Setting your fat intake by hormone anxiety is how people end up eating 40 percent of their calories from fat while wondering why their training feels flat.

The reasons the floor actually holds up

There are better arguments, and they are duller.

The first is absorption. Some dietary fat in a vegetable meal improves the absorption of carotenoids such as alpha-carotene, beta-carotene and lycopene. Brown and colleagues (2004) fed seven subjects identical vegetable salads with dressings containing 0, 6 or 28 grams of canola oil. With the fat-free dressing, the appearance of those three carotenoids in chylomicrons was negligible. Not reduced. Negligible. Absorption rose with the reduced-fat dressing and was higher again with the full-fat one. Seven people and a single meal each is a small study, so hold the exact numbers loosely, and note that this was carotenoids from a salad rather than every fat-soluble nutrient. A fat-free dressing can therefore reduce carotenoid absorption when the salad contains no other meaningful source of fat.

The second is that the food gets harder to arrange. At lower fat intakes, fitting normal portions of eggs, salmon, nuts, oils and cheese becomes increasingly difficult, especially when total calories are already low. None of it is off the table, but the lower you go, the more the day has to be planned around those foods rather than including them casually. Adherence is the variable that actually decides whether a diet works.

The third is palatability. For many people some fat improves a meal's palatability and satisfaction, and very low-fat meals can feel harder to sustain. That is not because fat is uniquely satiating. The research does not support that, and per calorie fat is generally less satiating than protein or carbohydrate. It is an adherence argument, not a physiological one.

What happens at the top end

The ceiling matters less, but it is real. At fixed calories and protein, additional fat leaves less room for carbohydrate. If that pushes carbohydrate below what your training requires, session quality can suffer. Note the conditions on that, though: what matters is your absolute carbohydrate intake, not the percentage on its own. A higher-fat split at generous calories can still leave plenty of carbohydrate, and the same split at 1,800 calories may not. Where those carbs sit around training matters less than most people think, but having enough of them matters quite a lot.

Setting your number

Here is how I do it, in order. The rule below is how I operationalize that uncertainty. It is a practical starting point, not an official consensus threshold.

Set protein first, from bodyweight, per the protein numbers. Then set a fat floor rather than a fat target: 20 percent of your calories, or 0.5 grams per kilogram of bodyweight, whichever of the two is higher. That is the number you defend. Then everything left over goes to carbs.

I am not the only one who has landed there. The ICECAP trial protocol (Peos et al., 2018) used the same operational floor in its diet prescription: 20 percent of calories from fat unless that fell below 0.5 g/kg, in which case 0.5 g/kg became the minimum. That is one research group's operational choice for a dieting athlete population rather than a consensus position, which is about the weight it deserves.

Worked example. An 80 kg lifter eating 2,400 calories. Protein at 1.8 g/kg is 144 g, which is 576 calories. The fat floor by bodyweight is 40 g; the floor by percentage is 20 percent of 2,400, or 480 calories, which is 53 g. Take the higher one, so fat sits at 53 g and 480 calories. That leaves 1,344 calories, or about 336 g of carbs.

Above the floor, the split is preference. Some people feel better with fat at 30 percent and fewer carbs. Some train better with the reverse. Neither is wrong, and the honest reading of the Delany review is that the expert groups do not have a strong opinion on where you land inside the range either.

The number that changes all of this is your calorie total, and that rests on a maintenance estimate that is rougher than it looks. What your calculator is actually estimating is worth understanding before you treat any of these grams as precise. The macro calculator gives you a starting point to adjust from over a few weeks, not a prescription.

The short version

Fat is not a lever you pull for a result. It is a floor you stop cutting into. Set protein from bodyweight, keep fat at the higher of about 0.5 g/kg or 20 percent of calories, give the rest to carbs, and stop worrying about your hormones at 25 percent fat, because the evidence that they are in trouble is genuinely mixed.

If you have been dieting a while and have never checked what your fat intake has drifted to, go and look. It is usually lower than you think.