How Much Do You Sweat?

A simple guide for marathon and ultramarathon runners

You can spend a small fortune on hydration products, electrolyte mixes and sweat tests, but one of the most useful numbers for an endurance athlete can be measured at home with a bathroom scale:

Your sweat rate.

Your sweat rate tells you approximately how much fluid your body loses during an hour of running.

It matters because runners are remarkably different. One athlete may lose half a litre of sweat per hour while another loses well over two litres under the same general conditions.

This guide is built around three important sports-science papers:

• Sawka et al., 2007, "Exercise and Fluid Replacement", which established practical guidance for individualising hydration during exercise.
• Baker, 2017, "Sweating Rate and Sweat Sodium Concentration in Athletes", which reviewed how sweat should be measured and why results vary so much between athletes and conditions.
• Barnes et al., 2019, "Normative Data for Sweating Rate, Sweat Sodium Concentration, and Sweat Sodium Loss in Athletes", which analysed sweat data from 1,303 athletes and showed just how large the differences can be between individuals and sports.

The practical message from all three is surprisingly simple:

Don't guess your sweat rate from what another runner drinks. Measure your own.

Option 1: The bathroom-scale method

For most runners, this is the best place to start.

Baker's 2017 review describes the change in body mass before and after exercise as the simplest and most accurate practical way to estimate whole-body sweat rate. Barnes and colleagues used essentially the same approach in their large athlete dataset.

You need:

• a reasonably accurate bathroom scale
• your running watch or phone for duration
• your water bottles
• ideally a measuring jug or bottles with known volumes

That's it.

Before your run

Go to the toilet first.

Then weigh yourself wearing as little as practical. Ideally, weigh yourself nude. Wet running clothes can hold quite a lot of sweat and make your finishing weight misleading.

Write the number down.

For example:

Starting weight: 72.4 kg

Now note how much fluid you take with you.

Suppose you start with two 500 mL bottles.

Do your run

For your first test, choose a normal training session of roughly 60–120 minutes.

Run at an effort that represents something you actually want to understand.

For example:

• easy long-run pace
• marathon effort
• steady ultramarathon effort
• hot-weather training

Don't deliberately drink more or less than usual simply because you're doing a test.

Record how much you actually drink.

After your run

If possible, towel yourself dry, remove wet clothing and weigh yourself again.

Suppose the scale now says:

Finishing weight: 71.8 kg

You therefore lost:

0.6 kg of body weight

For practical field testing, think of:

1 kg of weight loss ≈ 1 litre of fluid

So you lost approximately 600 mL from your body.

But that's not the entire story.

Imagine you also drank 700 mL during the run.

That 700 mL went into your body and helped disguise some of the sweat loss on the scale.

So:

600 mL lost from your body + 700 mL you drank = approximately 1.3 litres of sweat

If the run lasted exactly one hour:

Your sweat rate was approximately 1.3 L/hour.

If it lasted two hours, divide the total by two:

1.3 L ÷ 2 h = 0.65 L/hour.

That's really all the calculation most runners need.

One extra complication: going to the toilet

If you urinate during the test, make a rough note of how much.

Urine leaving your body reduces your finishing weight, but it obviously isn't sweat.

For example:

You calculate a total apparent loss of 1.5 litres, but you urinated approximately 300 mL.

Your estimated sweat loss is therefore closer to:

1.5 L - 0.3 L = 1.2 L.

For a simple 60–90 minute training test, choosing a session where you don't need a bathroom stop makes life considerably easier.

A complete example

Anna is preparing for a marathon.

She weighs herself immediately before a 90-minute marathon-pace run:

Before: 64.8 kg

After the run:

After: 63.9 kg

She therefore lost:

0.9 kg ≈ 900 mL

During the run she drank one 600 mL bottle.

So her estimated sweat loss was:

900 mL + 600 mL = 1,500 mL

The run lasted 1.5 hours.

So:

1,500 mL ÷ 1.5 hours = 1,000 mL/hour

Anna's sweat rate during that run was therefore approximately:

1.0 L/hour

Much easier than it sounds.

Don't measure it once and tattoo the number onto your hydration plan

Your sweat rate is not a fixed personal constant.

You don't have one magical number called "my sweat rate."

Baker's review shows that sweat rate changes with factors including exercise intensity, temperature, humidity, clothing, acclimation, body size and other individual characteristics. Even under controlled conditions, some day-to-day variation is normal.

Barnes and colleagues found exactly this sort of variability across 1,303 athletes. Endurance athletes averaged around 1.28 L/hour, but the standard deviation was 0.57 L/hour, meaning there was enormous variation around that average.

So knowing that "endurance athletes average about 1.3 L/hour" is interesting.

Knowing your sweat rate is useful.

Better still, know several versions of your sweat rate.

Build your own sweat-rate map

For marathon and ultramarathon training, repeat the scale test under different conditions.

You might eventually have something that looks like this:

Cool easy run:

0.6 L/hour

Cool marathon-pace run:

0.9 L/hour

Warm easy run:

1.0 L/hour

Warm marathon effort:

1.3 L/hour

Very hot long run:

1.6 L/hour

Now you have something far more useful than a generic hydration recommendation.

You have a rough map of how your body behaves.

This also explains why testing at easy pace in March may poorly predict what happens during a hard marathon in July.

Baker specifically recommends testing under conditions representative of the athlete's training or competition, including environment and exercise intensity.

Marathon runners: test race-like conditions

If you're preparing for a marathon, perform several tests during training sessions that resemble race day.

Try to match:

• expected temperature
• expected humidity if possible
• marathon pace or effort
• similar clothing
• similar time of day
• similar hydration before starting

You don't need laboratory precision.

You are looking for a useful range.

For example, after several tests you may discover:

My marathon-effort sweat rate in moderate weather is usually around 0.9–1.1 L/hour.

That's a much better planning tool than declaring:

My sweat rate is exactly 1.03 L/hour.

Ultramarathon runners need to be a little more careful

With ultramarathons, simply weighing yourself before and after a six- or ten-hour run becomes less accurate.

Baker points out that during very prolonged exercise, body mass also changes because of things other than sweat, including respiratory water loss and the mass lost through burning carbohydrate and fat. These effects become increasingly relevant beyond roughly 2–3 hours.

There's another problem.

During an ultra you may be:

• eating
• drinking
• urinating
• changing clothing
• carrying water
• getting soaked by rain
• pouring water over yourself
• crossing rivers

At some point the bathroom-scale experiment becomes an accounting department.

For ultramarathon planning, it is often more useful to perform several controlled 60–120 minute sweat-rate tests at different intensities and temperatures rather than trying to calculate sweat rate from one eight-hour adventure.

The simplest approach: scale only

If you don't care about gadgets, this is the approach I'd recommend.

1. Weigh yourself before running.

Ideally after going to the toilet and with minimal clothing.

2. Run for around 60–120 minutes.

Choose conditions and effort relevant to your race.

3. Keep track of what you drink.

Bottle markings are accurate enough for practical use.

4. Weigh yourself again.

Dry yourself first and avoid weighing soaked clothes.

5. Add together your weight loss and what you drank.

If you lost 0.8 kg and drank 0.5 L, you probably sweated approximately 1.3 L.

6. Divide by the number of hours you ran.

That's your approximate litres of sweat per hour.

7. Repeat the test.

Do it in cool weather, hot weather and at different race-relevant intensities.

The repeated measurements are more valuable than squeezing another decimal place out of one test.

The tech-assisted approach

If you enjoy data, you can build a much richer picture.

Modern sweat-sensing patches and wearable systems can estimate local sweat rate, and some can also estimate electrolytes such as sodium.

That sounds wonderful, but there's an important distinction:

A sensor attached to your arm measures what is happening at that piece of skin.

What you usually want for hydration planning is:

whole-body sweat loss.

Baker's review spends considerable attention on this problem. Local sweat measurements can be useful, but sweat rate and sweat composition differ depending on where and how they are measured.

So if you use wearable sweat technology, don't throw your bathroom scale into retirement.

Use the two together.

A good tech setup

For a race-specific sweat test, you could record:

Body weight before and after

This remains your anchor for whole-body fluid loss.

Wearable sweat sensor or patch

This can show how local sweating develops through the session and, depending on the device, may estimate sweat sodium.

Running watch

Record heart rate, pace, power or another measure of intensity.

Weather data

Record temperature and humidity.

Now instead of knowing only:

"I sweat 1.1 L/hour."

you can learn:

"At easy effort around 15°C, I sweat around 0.7 L/hour."

"At marathon effort around 20°C, I'm closer to 1.0 L/hour."

"At similar effort around 30°C, I'm around 1.4 L/hour."

That is where technology becomes genuinely useful.

It helps you build a personal sweat profile, rather than simply producing another number for your training dashboard.

Don't assume a wearable is automatically more accurate

The shiny sensor is not necessarily superior to the humble scale.

Recent research on wearable sweat-loss technology has found that devices can correlate reasonably with actual sweat loss while still producing substantial individual errors. One 2025 evaluation reported mean absolute percentage errors of roughly 25–33% for the wearable devices tested and concluded that traditional measurement methods remain important when accuracy matters.

For practical endurance training:

Use technology to add information.

Use body-mass change as your reality check.

What is a "normal" sweat rate?

There is a huge range.

Baker's review reports that whole-body sweat rates in athletes commonly fall around 0.5–2.0 L/hour, although rates above 3 L/hour can occur in unusual circumstances.

Barnes's dataset illustrates the same variability. Among endurance athletes, the average was approximately:

1.28 L/hour

with a large variation between individuals.

So if you measure 0.7 L/hour, that doesn't mean you're "not sweating enough."

If you measure 1.8 L/hour, that doesn't automatically mean something is wrong either.

The purpose of measuring isn't to win the Sweat Olympics.

It's to understand what your body is losing so you can make better decisions.

Sweat rate is only half of the hydration puzzle

One final distinction matters enormously.

Sweat rate tells you how much water you're losing.

It does not tell you how much sodium you're losing.

Two runners can each sweat:

1 litre/hour

while losing very different amounts of sodium because their sweat sodium concentrations are different.

This is one of the central points of Baker's review and the Barnes dataset. Barnes calculated sodium-loss rate by combining an athlete's sweat rate with their sweat sodium concentration.

So think of sweat testing as two separate questions:

Question 1: How much do I sweat?

Measured reasonably well at home with a scale.

Question 2: How salty is my sweat?

This requires some form of sweat-sodium measurement if you want an individualized estimate.

Combine those two and you can begin estimating your sodium loss per hour.

That comes next.

The takeaway

You don't need a laboratory to learn something valuable about your hydration.

Start with a scale.

Measure yourself before and after a representative run.

Keep track of what you drink.

Repeat the experiment in different weather and at different effort levels.

Rather than searching for one perfect sweat-rate number, build a small collection of measurements that represents the conditions you actually race in.

The science from Sawka 2007, Baker 2017 and Barnes 2019 points in the same direction:

Hydration should be individualised because sweating is individual.

For most marathon and ultramarathon runners, a bathroom scale, a bottle and a few carefully chosen training runs are enough to begin discovering just how individual it is.