Back

Hydration series - Part 2 of 4

You Know Your Sweat Rate. Now How Much Should You Actually Drink?

A practical hydration guide for marathon and ultramarathon runners

You've done the sweat test.

You weighed yourself before a representative run, kept track of what you drank, weighed yourself afterwards and discovered something like:

My sweat rate is about 1.2 litres per hour.

Great.

Now comes the question that causes far more confusion:

Does that mean I should drink 1.2 litres every hour?

Usually, no.

Your sweat rate tells you how quickly fluid is leaving your body. It does not automatically tell you how quickly fluid needs to go back in.

That distinction is central to the hydration guidance from Sawka et al. 2007, the sweat-testing work reviewed by Baker 2017, and the large athlete dataset published by Barnes et al. 2019. Sawka specifically recommends individualized fluid plans because sweat losses vary so widely between athletes, while Baker and Barnes show just how large that variability really is.

The goal isn't:

Sweat 1 litre → drink 1 litre.

The goal is:

Drink enough to avoid becoming excessively dehydrated, but not so much that you overdrink.

That leaves some room in the middle.

And that middle is where a good race hydration strategy lives.


First: think of sweat rate as your loss rate

Imagine you measure a sweat rate of:

1.0 L/hour

If you run for four hours under similar conditions, your body could lose roughly:

4 litres of sweat.

But that does not mean you need to drink four litres during the race.

Your body contains a substantial reservoir of water, and some loss of body water during endurance exercise is normal.

Sawka's 2007 ACSM position stand proposed avoiding excessive dehydration, commonly defined there as losing more than about 2% of body mass from water deficit.

So instead of asking:

"How do I replace every drop of sweat?"

a more useful question is:

"How much of my sweat loss do I need to replace during this particular race?"

Your body weight gives you a useful buffer

Suppose you weigh:

70 kg

Two percent of your body weight is:

1.4 kg

For practical hydration calculations, think of that as roughly:

1.4 litres

So, using the traditional Sawka approach, a 70 kg runner can lose roughly 1.4 litres of body water before crossing the 2% body-mass-loss mark.

That gives us something useful.

You don't necessarily have to replace the first 1.4 litres of sweat immediately.

You have some room between being perfectly hydrated and becoming substantially dehydrated.


Let's put this into a marathon

Imagine David weighs 70 kg.

From several training tests, he knows that in conditions similar to race day he sweats approximately:

1.0 L/hour

He expects to run the marathon in four hours.

Without drinking anything, he might lose roughly:

4 litres

during the race.

If he wanted to finish no more than around 1.4 litres down, he would need to replace approximately:

4.0 L lost - 1.4 L allowable deficit = about 2.6 L

Spread across four hours:

about 650 mL/hour

Notice what happened.

His sweat rate is:

1,000 mL/hour

but his drinking rate might be closer to:

650 mL/hour

He doesn't have to chase every bead of sweat with a mouthful of water.


This is why sweat rate is not your drinking rate

A runner sweating:

1.2 L/hour

might drink:

600 mL/hour

or:

750 mL/hour

or:

900 mL/hour

depending on:

  • body size
  • race duration
  • temperature
  • intensity
  • how much dehydration they tolerate
  • gastrointestinal tolerance
  • access to aid stations
  • thirst
  • previous experience

There isn't a universal percentage of sweat loss that every athlete needs to replace.

Baker specifically recommends using ranges rather than pretending sweat measurements give an exact prescription, because even carefully performed sweat-rate tests show natural day-to-day variation.

Your hydration target should therefore look more like:

"Approximately 500–700 mL/hour in these conditions"

than:

"I must drink exactly 643 mL every hour."

Your stomach does not own a laboratory pipette.


The simplest practical rule

Once you've measured your sweat rate, think of it as a boundary, not an order.

If you normally sweat around:

0.8 L/hour

you probably shouldn't construct a plan requiring you to drink 1.1 L/hour under those same conditions.

You would be putting fluid into your body faster than you're losing it.

Over several hours, that can produce weight gain and increase the risk of exercise-associated hyponatremia, a potentially dangerous fall in blood sodium concentration.

Later consensus guidance on exercise-associated hyponatremia emphasizes that the major preventable risk is overdrinking, particularly drinking fluid in excess of losses. Athletes should generally not gain weight during prolonged exercise.

So one extremely useful rule is:

Don't try to finish an endurance race heavier than you started.


What about the famous 400–800 mL per hour recommendation?

Sawka's 2007 guidance is often summarized as a general drinking range of approximately:

400–800 mL/hour

for many marathon-type situations.

But Sawka did not mean:

Everybody should drink 600 mL/hour.

The paper specifically discusses adjusting intake according to factors such as body size, running speed and environmental conditions. Faster, larger runners in warmer conditions are generally toward the higher end, while smaller, slower runners in cooler weather may need considerably less.

This makes much more sense once you've measured your sweat rate.

Imagine three runners.

RunnerMeasured sweat rate600 mL/h would replace
Runner A0.6 L/h~100% of sweat
Runner B1.0 L/h~60%
Runner C1.5 L/h~40%

All three could follow the identical advice:

"Drink 600 mL/hour."

Yet physiologically they would be doing three very different things.

This is precisely why individualized sweat testing matters.

Barnes et al. found an average sweat rate of around 1.28 L/hour in endurance athletes, but with huge variation between individuals.

Population averages tell you what runners tend to do.

Your sweat test tells you what you tend to do.


A very simple way to build your starting drinking plan

You need only three pieces of information:

1. Your body weight

2. Your sweat rate under race-like conditions

3. How long you expect to be racing

Then ask:

How much fluid will I probably lose?

If your sweat rate is 0.9 L/hour and your marathon will take four hours:

You're looking at approximately 3.6 litres of sweat.

How much body-mass loss would 2% represent?

If you weigh 65 kg:

2% is about 1.3 kg, or roughly 1.3 litres for this simple field estimate.

How much would I therefore need to replace to stay around that traditional 2% boundary?

You might aim to replace roughly:

3.6 - 1.3 = 2.3 litres

over four hours.

That's around:

575 mL/hour

Don't obsess over the last 25 mL.

Call it:

roughly 500–650 mL/hour

and test that range in training.

That is a hydration plan.


But 2% isn't a magical cliff

It's worth understanding what the number means.

If you finish at 2.1% body-mass loss, your body does not suddenly ring an alarm bell.

Likewise, maintaining exactly 0.0% body-mass loss doesn't automatically produce the best performance.

The 2% figure is a useful traditional guideline from the Sawka position stand, not a precision boundary separating "hydrated" from "dehydrated."

In real endurance racing, some athletes finish successfully with more than 2% body-mass loss.

The important principle is to avoid the two extremes:

Too little fluid → increasingly large fluid deficit

and

Too much fluid → overhydration

For marathoners and especially ultrarunners, the second one deserves more attention than many athletes realize.


The dangerous mistake: "I must replace 100% of my sweat"

Suppose you sweat:

1.4 L/hour

You decide:

"Excellent. I'll drink 1.4 litres every hour."

Now imagine your actual race conditions are cooler than the day you performed your sweat test.

Your real sweat rate might now be only:

1.0 L/hour

But you're still drinking:

1.4 L/hour

Every hour you're putting approximately 400 mL more fluid into your body than you're losing through sweat, before considering urine and other losses.

Over five hours that discrepancy can become substantial.

This is one reason Baker recommends testing under conditions representative of actual competition and repeating measurements across different conditions. Sweat rate changes with temperature, intensity, acclimation, clothing and other factors.

Your hot-weather sweat rate should not automatically become your cool-weather drinking plan.


The ultra changes the game

A marathon lasts several hours.

An ultramarathon might last:

6 hours.

12 hours.

24 hours.

Or considerably longer.

That means small mistakes have time to grow teeth.

Suppose your true fluid requirement is around:

600 mL/hour

but you force yourself to drink:

800 mL/hour

The difference is only:

200 mL/hour.

It doesn't sound dramatic.

But over 12 hours:

200 mL × 12 = 2.4 litres

That's potentially a very large excess.

The opposite is also true.

If you're losing 1.2 L/hour in hot conditions but consistently manage only 300 mL/hour, the deficit can accumulate rapidly.

This is why ultramarathon hydration shouldn't be built around a generic bottle-per-hour rule.


Ultrarunners have one useful advantage: checkpoints

In some long races you can monitor body mass during the event.

This isn't perfect because prolonged exercise changes body mass for reasons beyond simple water loss, something Baker discusses particularly for exercise lasting more than about 2–3 hours.

But one observation is still particularly useful:

Weight gain during a long race is a warning sign.

If you started at 70 kg and several hours later you're 71 kg, you haven't somehow created new muscle during the race.

You've probably retained more fluid than you've lost in net terms.

Guidelines addressing exercise-associated hyponatremia specifically recommend reducing fluid intake when an athlete is gaining weight during prolonged exercise.


What role should thirst play?

Quite a big one.

A sweat-rate plan should not turn you into a drinking robot.

Your pre-race sweat testing tells you the general neighbourhood of your requirements.

Thirst gives you useful real-time information.

Modern guidance on preventing exercise-associated hyponatremia considers drinking according to thirst a safe strategy for most endurance athletes, while also recognizing that unusually large and rapid fluid losses may sometimes require more deliberate planning.

The two ideas don't have to fight.

Think of it this way:

Sweat testing gives you the map.

Thirst helps you steer.

If your training data suggest:

500–700 mL/hour

and you're comfortably drinking somewhere in that region according to thirst, that's very different from ignoring thirst and forcing down 1.2 L/hour because a spreadsheet told you to.


Your stomach gets a vote

There is another limit that runners discover rather quickly.

You may calculate that replacing a large proportion of your sweat loss requires:

900 mL/hour

But if drinking 900 mL/hour leaves you bloated, nauseated and sloshing down the road, it's not a useful race plan.

Fluid intake has to be practiced just like carbohydrate intake.

During long runs, test:

400 mL/hour

then perhaps:

500 mL/hour

then:

600 mL/hour

and learn what feels comfortable under realistic racing conditions.

Your hydration plan has to survive contact with your gastrointestinal system.


Marathon example

Meet Sarah.

Body weight: 60 kg

Typical marathon-effort sweat rate: 0.8 L/hour

Expected marathon time: 4 hours

Over four hours she might sweat approximately:

3.2 litres

Two percent of her starting body mass is approximately:

1.2 kg

So a traditional Sawka-style plan designed to finish around or below 2% body-mass loss might involve replacing approximately:

3.2 - 1.2 = 2.0 litres

over four hours.

That's:

about 500 mL/hour

Sarah doesn't need to drink her full 800 mL/hour sweat rate.

She can tolerate some fluid deficit while running.

So she might begin training with a practical target around:

450–600 mL/hour

and adjust according to weather, thirst and experience.


Hot marathon example

Now imagine the same runner tests again during warm conditions.

Instead of:

0.8 L/hour

she measures:

1.2 L/hour

Over four hours that's approximately:

4.8 litres of sweat

Suddenly her cool-weather drinking plan of 500 mL/hour may leave a much larger deficit.

Her hydration strategy probably needs to move upward.

This is exactly why Baker recommends repeating sweat tests under different environmental conditions and exercise intensities.

There isn't one Sarah hydration plan.

There is:

Sarah in cool weather.

Sarah in moderate weather.

Sarah in hot weather.

Same runner. Different plumbing.


A practical race-day system

Instead of memorizing equations, build three numbers from training.

Cool conditions

Example:

Sweat rate: 0.7 L/h

Typical comfortable drinking: 0.4–0.5 L/h

Moderate conditions

Sweat rate: 1.0 L/h

Typical comfortable drinking: 0.5–0.7 L/h

Hot conditions

Sweat rate: 1.4 L/h

Typical comfortable drinking: 0.7–0.9 L/h

These are only example numbers, not recommendations.

Your numbers may look completely different.

The point is to build a personal hydration map rather than a single hydration rule.


Then turn litres per hour into something useful

"Drink 600 mL/hour" isn't terribly useful when you're approaching an aid station at kilometre 27.

Translate it into things you can see.

For example:

Target: approximately 600 mL/hour

That could mean:

One 500 mL soft flask per hour plus a few drinks at aid stations.

Or:

About 150 mL every 15 minutes.

Or:

Two 300 mL cups/bottles per hour.

Or for an ultra:

One 600 mL flask between aid stations expected to take approximately an hour.

Don't build a hydration strategy in laboratory units.

Build it in bottles, flasks and aid stations.


Three rules worth remembering

If everything else disappears from your brain at kilometre 35, remember these:

1. Your sweat rate is not automatically your drinking rate.

You generally don't need to replace every millilitre of sweat while you're running.

2. Don't consistently drink more than you're losing.

In prolonged endurance exercise, overdrinking can lead to fluid retention and exercise-associated hyponatremia. Sodium supplements do not make excessive fluid intake safe.

3. Test the plan in training.

Baker and Barnes make clear that sweat losses vary dramatically between athletes and across conditions. Your race plan should therefore come from your measurements under representative conditions, not somebody else's hydration strategy.


So what should you actually drink?

After measuring your sweat rate, don't simply copy that number into your hydration plan.

Instead:

Know approximately how much you're losing.

Allow for some normal body-water loss.

Avoid excessive dehydration.

Never deliberately drink yourself into weight gain.

Let thirst help fine-tune the plan.

Adjust for heat and effort.

Practice the amount your stomach can comfortably handle.

And think in ranges rather than exact numbers.

A runner who knows:

"In these conditions I lose around 1.0 L/hour, and 500–700 mL/hour works well for me"

has a far more useful hydration strategy than a runner who simply reads:

"Drink 750 mL every hour."

Where the three research papers fit

Sawka et al. 2007 gives us the practical hydration framework: begin exercise appropriately hydrated, avoid excessive dehydration, avoid major electrolyte disturbances, and individualize fluid replacement rather than applying one number to everyone.

Baker 2017 explains why individualization matters. Sweat rate varies substantially with the athlete, intensity, environment, acclimation and other conditions. It also tells us how to measure sweat rate properly and reminds us not to pretend a field measurement is more precise than it really is.

Barnes et al. 2019 shows this variability in the real world. In more than 1,300 athletes, endurance athletes had some of the highest sweat rates and sodium-loss rates, but the variation among individuals was enormous.

Together they give us a useful sequence:

Measure → Estimate → Test → Adjust

Measure your sweat rate.

Estimate a reasonable drinking range.

Test that range during training.

Adjust it for weather, effort, duration and thirst.

That's a much better hydration strategy than attempting to replace every drop that leaves your skin.


One more piece is missing

Knowing how much fluid to drink still doesn't tell you how much sodium to take.

That's because:

Sweat rate = how much fluid you're losing

while:

Sweat sodium concentration = how salty that fluid is

Two runners can both lose 1 litre of sweat per hour while losing dramatically different amounts of sodium.

That's where the next part of the puzzle begins:

You Know Your Sweat Rate. Now How Much Sodium Do You Actually Need?