How Much Food to Carry: The Number That Matters Is Calories per Gram

Start with an equation, not a rule of thumb

The Pandolf equation was published in 1977 by Pandolf, Givoni and Goldman at the US Army Research Institute of Environmental Medicine, and it is still the standard tool for predicting the metabolic cost of walking with a load. It adds three terms: a standing cost proportional to body mass, a penalty for carrying a load that grows with the square of the load-to-body ratio, and a walking term that scales with total mass, the square of speed, the grade, and a dimensionless terrain factor. The terrain factor is what makes it useful off a road. Blacktop is 1.0, a dirt road 1.1, light brush 1.2, heavy brush 1.5, a swampy bog 1.8, loose sand 2.1, and soft snow runs from 2.5 to 4.1 depending on how deep you sink.

It assumes a lot, and the assumptions are worth naming. It was fitted to soldiers walking on a treadmill, so it describes steady-state walking on uniform ground rather than scrambling, step-ups or route-finding. It was validated for loads up to about 40 kg, speeds up to about 6 km/h and grades from level to ten per cent, so it is out of range on a steep pass. It has no downhill term at all — descending grades need a separate correction, because walking downhill costs less than level walking up to a point and then costs more. And it is not conservative in the way you might hope: a 2017 study of contemporary military load carriage found it under-predicted measured metabolic rate, with a mean bias of about 125 watts. In planning terms, a number this equation gives you is a floor, not a ceiling.

There is a simpler alternative, the ACSM walking equation, which estimates oxygen uptake from speed and grade alone. It is easier to use and it has one fatal gap for this problem: it has no load term at all, so it cannot answer the question a backpacker is actually asking. Pandolf is the right tool here precisely because the pack is one of its inputs.

One day, worked through

A 155-pound hiker carrying a 33-pound pack, walking eight hours at 2.8 mph over reasonable trail — terrain factor 1.2, the light-brush value — comes out at about 354 watts of metabolic rate, which is 304 calories an hour and 2,435 calories for the walking day. The other sixteen hours are not free: at a resting metabolic rate of about 1,650 calories a day, sixteen hours of camp chores and sleep costs roughly 1,320. Total, about 3,750 calories, which for planning purposes is 3,700 and a bit.

Two things move that number a lot, and it is worth knowing which. Terrain is one: the same eight-hour day on blacktop is about 3,460 calories, on a dirt road 3,600, in heavy brush 4,150, and in loose sand nearly 4,970. Between a forest road and a dune field the day changes by 1,500 calories, which at a realistic energy density is nearly ten ounces of extra food. Grade is the other, and it is brutal: hold everything else fixed and put that walk on a five per cent grade and the metabolic rate jumps 63 per cent. Three graded hours in an eight-hour day push the total from 3,700 to about 4,300.

The binding constraint is mass, so plan in calories per ounce

Knowing you need 3,700 calories tells you nothing about what goes in the pack, because calories have no weight. What has weight is the food, and the exchange rate between them is the energy density. This is why experienced planners talk in calories per ounce and not in calories per day: the calorie figure is a requirement, but the density is a decision, and the decision is where all the leverage lives.

The table above runs the same 3,700-calorie day across the realistic range. At 115 calories per ounce, a day weighs two pounds and a seven-day carry weighs 14.1. At 155 calories per ounce, the same day weighs a pound and a half and the week weighs 10.4. Three and a half pounds of difference, without changing a single calorie of what you eat. For comparison, three and a half pounds is a two-person tent, or a warm sleeping bag's worth of down, or two litres of water.

Where does the density come from? From the Atwater factors, which the FAO sets out: fat yields about 9 calories per gram, carbohydrate and protein about 4 each, and water yields nothing at all. Solve for the fat share of a dry food mix and the arithmetic is unforgiving. To reach 4.5 calories per gram, roughly a tenth of the dry mass must be fat. To reach 5.0, a fifth. To reach 5.5, about 30 per cent. The theoretical ceiling is 9 calories per gram — pure fat, in which case the whole day weighs 411 grams — and nobody eats that. Which means the practical range is not a preference: it is a line drawn between what physics allows and what a person will still swallow on the fifth evening.

The ceiling: you cannot absorb an arbitrary amount

The arithmetic so far assumes you eat what you carry and absorb what you eat. Over a few days that holds. Over weeks it does not, and the reason is measurable. Thurber and colleagues assembled energy expenditure data across human endurance events lasting from half a day to more than 250 days, and added their own measurements from runners crossing a continent. The pattern is a curve: the longer the effort, the lower the sustainable multiple of basal metabolic rate, plateauing below three times basal. Combining that with overfeeding studies, they identified an alimentary supply limit around 2.5 times basal metabolic rate — a ceiling on how much energy the gut can actually deliver, above which the difference has to come out of the body's own stores.

Put our hiker's numbers against that. His basal rate is about 1,650 calories, so his alimentary ceiling is about 4,100 calories a day. The flat day we calculated costs 3,700, which is 2.27 times basal — comfortably inside. The graded day costs 4,300, which is 2.63 times basal, and that is over the line. On the graded itinerary he will lose weight no matter how carefully he packs, because the limit is not what he brought, it is what he can process. The right planning response is not more food. It is to expect the deficit, keep it small, and accept that a long hard route is a slow calorie draw-down rather than a break-even proposition.

Water is heavier than food, and it is a different problem

A litre of water weighs a kilogram — about 2.2 pounds — and unlike food it is entirely indigestible mass with zero energy density. The joint position statement of the Academy of Nutrition and Dietetics, Dietitians of Canada and the ACSM gives sweat rates during exercise ranging from 0.3 to 2.4 litres an hour depending on intensity, duration, fitness, acclimatisation and conditions, and notes that a kilogram of body mass lost is approximately a litre of sweat. Take a middling 0.7 litres an hour across six hours of walking and the day's fluid loss is 4.2 litres, weighing 4.2 kilograms. That is 5.7 times the mass of the whole day's food at 5 calories per gram.

The saving grace is that water is a flow problem rather than a stock problem. Food has to be carried for the whole leg; water only has to be carried between sources. On a well-watered trail the correct answer is often to carry half a litre and drink at every stream, and on a dry stretch it is to carry the gap and nothing more. The same position statement notes that fluid deficits beyond about two per cent of body weight begin to compromise aerobic performance and cognition, particularly in the heat, and that an intake of 0.4 to 0.8 litres an hour suits most people and most events. Between those two numbers sits the whole art of dry-section planning: know where the next source is, know what the gap costs in kilograms, and do not carry a litre past a spring.

Resupply arithmetic, and why food does not pay for itself

Food mass falls linearly as you eat it, so a carry's average weight is half its starting weight. A seven-day carry at 1.6 pounds a day starts at 11.4 pounds and averages 5.7. Split the same week with one resupply and the carry starts at 5.7 pounds and averages 2.9 — the peak halves and so does the mean. That is the entire case for resupply in one line, and it is a strong one, because the peak is what your hips feel on the first morning out of town and the mean is what you carry for the rest of it.

There is a tempting second argument that turns out to be weak, and it is worth killing. Carrying less should cost less energy, so a lighter pack should reduce the food requirement, which should reduce the pack weight again. Run it. Halving the load from 33 to 16.5 pounds drops the metabolic rate from 354 to 326 watts, saving about 187 calories over an eight-hour day. At 140 calories per ounce that is 1.3 ounces of food. The quadratic load-penalty term does behave dramatically — it falls by a factor of 4.4 — but it was only about 8 watts out of 354 to start with, so the drama is invisible in the total. Lighten your pack because carrying less is pleasanter and safer, not because the food will pay you back. It will not.

Frequently asked questions

How many calories a day should I plan for?

Do not use a single number, use the equation, because the honest range is enormous. The same person walking the same eight hours costs about 3,460 calories on a road and about 4,970 in loose sand, and putting three hours of five per cent grade in the day adds another 600. What moves the answer, in order of importance: grade, terrain, speed, load and your own body mass. A reasonable planning habit for a trail day with a moderate pack is somewhere around 3,500 to 4,000 calories, but the moment your route is steep, soft or long you should compute it rather than assume it, and remember that this equation has been shown to under-predict rather than over-predict.

Is there a target calories-per-ounce figure I should aim for?

The traditional target is 125 calories per ounce, and the arithmetic explains why it endures: it is reachable with ordinary food, needs only about a tenth of the dry mass as fat, and lands a seven-day carry at 13 pounds. Pushing to 155 saves another 2.6 pounds over the week but demands roughly 30 per cent of the dry mass as fat, which is a real palatability cost by the fifth evening. Below about 100 the food is carrying water you do not need. Where you sit in that band should follow trip length: on a two-night trip the density hardly matters and you should eat well; on a nine-day carry it is the single biggest lever on your pack weight.

Why does my appetite disappear for the first two days and then explode?

The pattern is common enough that experienced walkers plan around it, though the mechanisms are not fully settled and anyone claiming a tidy explanation is going beyond the evidence. What is not in dispute is the planning consequence, and it is useful: a deficit accumulated on days one and two has to be repaid later, and later is when your pack is lightest and your resupply may already be behind you. The practical response is to front-load palatability rather than calories — put the food you will actually want on the first days — and to keep the densest, most calorie-heavy items for the back half of the leg, when appetite has caught up and every gram saved earlier has already been carried.

Should I plan to lose weight on a long trip?

On a hard route you should expect it, because the alimentary ceiling makes it arithmetic rather than willpower. Sustained energy supply plateaus around 2.5 times basal metabolic rate, so once the day's expenditure exceeds that, the shortfall comes out of stored energy whatever you pack. Our hiker's graded day at 2.63 times basal is already past it. The right response is to keep the deficit modest and deliberate rather than to fight it: eat to appetite, favour density on the long carries, and treat a week of hard walking as a slow withdrawal rather than a balanced account. A deficit that becomes large is a different matter — it degrades judgement and cold tolerance, which on a remote route is a safety issue rather than a nutrition one.

How much water should I actually carry?

Carry the gap, not the day. Water is the only heavy item on your back that is replaceable en route, so the correct quantity is whatever gets you to the next reliable source with a margin, and nothing more. Sweat rates during exercise run from 0.3 to 2.4 litres an hour depending on heat, intensity and acclimatisation, and an intake of 0.4 to 0.8 litres an hour suits most people, so a three-hour dry stretch in warm weather is a two-litre problem and not a five-litre one. Weigh yourself before and after a long hot day if you want your own number: a kilogram lost is about a litre of sweat. And watch the threshold rather than the total — deficits beyond about two per cent of body weight start to cost you performance and clear thinking.