Measure Sweat Rate and Turn It Into a Drinking Plan

Whole-body sweat rate in Baker's 2016 cohort of 506 athletes ran from 0.26 to 5.73 L/h, with a mean of 1.21 L/h.¹ A spread that wide makes any borrowed drinking rule a guess. Your own number earns its keep once you convert it into a drink rate that keeps finish body-mass loss below 2% and never lets you finish heavier. Start hydrated; use 2% as a ceiling, not a target. Whatever the gut cannot empty gets replaced after the session. Forcing 100% replacement during it can overshoot your actual losses when pace or conditions change, which raises the risk of exercise-associated hyponatremia.¹⁶

A 55 kg runner at 0.60 L/h in a cool 10K can finish 0.9% lighter with no drinking. An 80 kg marathoner at 1.60 L/h in heat for 3.5 hours gets a 1.14 L/h drink rate from the 2% ceiling calculation before any event-urine term; choose a smaller finish-loss budget for a real plan, and many guts empty closer to 0.80 L/h. The field sweat-rate formula produces L/h. Convert it before you pack bottles.

Drink L/h ≈ sweat rate + expected event urine rate − (finish-loss budget × body mass ÷ session hours). Set the budget below 0.02, use an observed or rehearsed event-urine rate when available, floor the result at zero, then cap it at what the gut can empty.

For a session without urination, use zero for event urine. For a multi-hour event, include the urine rate from a race-like rehearsal when you have one.

01How to run a field sweat test and size its errors

Void first. Weigh nude before and after the session, or change into a known dry, pre-weighed kit before the post-weigh. Avoid food during the test so consumed mass does not distort the balance. Weigh every bottle. Collect any urine and weigh it. Towel dry before the post-weigh. The useful window is 45 to 90 minutes of the actual work. A 20 minute warmup mostly captures the lag before sweating reaches a steady rate, so it cannot stand in for the test.³⁴

The worked example in Baker's SSE 161 runs 82.5 kg pre, 81.0 kg post, 1.1 L drunk, 0.3 L urine, 1.75 h. Loss is 82.5 minus 81.0 plus 1.1 minus 0.3, which is 2.3 L. Divide by 1.75 h and you get 1.3 L/h.³ Treat 1 kg as 1 L in the field.

Ignore urine and the rate can run high, depending on urine volume relative to total sweat loss. In Cheuvront's 2002 study of eight women running 30 km, the error was 16 to 37%; it is zero when no urine is passed.⁵³ Finish in soaked clothes and trapped sweat hides 8 to 10% of the loss. Respiratory water and substrate mass look like extra sweat. Cheuvront puts that non-sweat mass near 0.20 g per kcal, about 5 to 15% of the scale change. Sixty minutes at 12 kcal/min is roughly 144 g that is not sweat.⁶

Skip that kcal correction on a 45 to 90 min easy or moderate test. Apply it for 2 to 3 hours, high intensity, or dry air. Day-to-day whole-body sweat rate moves about 5 to 7% when the method stays the same. Retest when weather, pace, or a heat block changes.⁴

02Convert L/h into a sub-2% finish budget

Two percent is the ceiling, not the target. Print that maximum-loss volume in liters before you pick a bottle, then choose a smaller budget when starting hydration or measurement uncertainty warrants it.

Subtract that ceiling from expected loss for the boundary calculation, spread the remainder across hours, and you have a drink rate. For 80 kg over 3.5 h at 1.60 L/h, gross loss is 5.60 L. The 2% ceiling is 1.60 L. Drink 4.00 L in 3.5 h, which is 1.14 L/h; choose a smaller loss budget for a real plan when conditions or starting hydration warrant it.

NATA 2017 uses about 2% body-mass loss as the performance guardrail, with heart rate up 3 to 5 bpm and core temperature up 0.15 to 0.20 °C per 1% lost.⁷ ACSM 2007 put typical sweating at 0.5 to 2.0 L/h. NATA's adult range runs 0.5 to 4.0 L/h. Four liters per hour is a recorded extreme. It is a terrible planning default.⁷⁸

Two percent is environment-dependent. Sawka, Cheuvront, and Kenefick found a 2% penalty in 0 of 2 cold studies at 2 to 10 °C, 4 of 7 temperate studies at 20 to 24 °C, and 8 of 9 warm-hot studies above 25 °C. Skin warmer than about 27 °C is where 2% starts to cost, then about 1.5% extra impairment per 1 °C of skin temperature.⁹ James 2017 blinded cyclists at about 2.4% versus 0.1% body-mass loss and still cut cycling work from 165 to 152 kJ, about 8%.¹⁰ Do not sell 2% to a cool-room lifter as a performance law. In an ultra, part of the scale drop is burned fuel and the water stored with glycogen, so a drop near 3% can overstate the real water deficit.⁶

If you have no test, ACSM's marathon starting band is 0.4 to 0.8 L/h.⁸ Once you have L/h, throw the band out and do bottle math.

Slot that rate into the aid-station script in the race-week nutrition plan. Carry it in a sports drink when the session also needs carbohydrate and sodium. Daily hydration intake and beverage choice are in the Complete Guide to Hydration.

03Why one treadmill number cannot plan every race

A cool treadmill L/h cannot set a hot marathon. Test in the three conditions you actually race or train in. Easy-cool. Race-pace heat. Indoor if the trainer, the track, or the gym is a real mode.

Heat acclimation changes both sides of the budget. Kirby 1986 had ten men train 10 days at 40 °C. Sweat rate rose, plasma volume rose 12%, and sweat sodium per liter fell.¹¹ Buono 2018, in four people, saw the local sweat-sodium curve drop about 45% at a given sweat rate across 7 days.¹⁸ Retest after a heat-acclimation block. The spring number is expired.

Indoor is its own condition. There is no honest indoor-plus-X% multiplier. A 70 kg athlete at 1.40 L/h for 90 min indoors loses 2.10 L. Use an illustrative 1.5% finish-loss budget of 1.05 L. Drink 1.05 L in 1.5 h, about 0.70 L/h, and replace the other 1.05 L after the session. The same athlete at 0.90 L/h outdoors in a breeze has a smaller hole. Measure both.

Barnes 2019 put adult means at 1.20 L/h and youth at 0.85, males at 1.18 and females at 0.82, football at 1.51 and baseball at 0.83.² Your teammate's bottle plan is a different physiology in a different shirt.

04How to use a regional sweat-sodium reading

Milligrams per hour equal mmol/L times 23 times liters per hour when the concentration represents whole-body sweat. The 36 mmol/L value in these datasets is a regional patch reading, so pairing it with 1.2 L/h calculates about 990 mg/h as an illustration, not a personal whole-body loss estimate.¹² Derive intake from the planned fluid volume and sodium concentration, then adjust for duration, conditions, and tolerance. Visible salt on a dark shirt is a crude flag. Fabric, evaporation, and session length all change the stain. The stain does not give you 50 versus 70 mmol/L.

The table below shows the arithmetic for regional readings; it does not convert a patch sample into a whole-body loss measurement.

ACSM's during-exercise sodium band is 20 to 30 mEq/L, 460 to 690 mg/L of drink.⁸ Thomas 2016 says take the sodium plan more seriously when sweat rate is above 1.2 L/h, sweat is salty, or the event runs past 2 hours.¹² Use the planned fluid volume and concentration to derive intake, then adjust for duration, conditions, and tolerance. Pair it with the electrolyte balance plan. Most amateurs miss sodium at the food level, which is the argument of The Amateur Athlete's Sodium Mistake. Sodium loading is a separate pre-race plasma-volume tool. Full replacement of sweat sodium during the session is not the target.

In Hoffman 2015, 93.9% of 161 km runners used sodium supplements, and hyponatremic finishers were no less likely to have taken them.¹³ Overdrinking still wins. Capsules do not cancel a rising scale.

05When the equation says drink less than you planned or more than your gut allows

Plug the 55 kg, 50 min, 0.60 L/h 10K into the equation. Session loss is 0.50 L. The 2% ceiling is 1.10 L. Finish mass loss is 0.9% with zero drinking. The equation returns a negative drink rate, which floors at zero. Carry a bottle for comfort if you want. The mass budget is already below the ceiling.

The 80 kg hot marathon goes the other way. At 1.60 L/h for 3.5 h, the 2% ceiling calculation yields 1.14 L/h before event urine; 0.5 L of event urine would add about 0.14 L/h. A smaller loss budget would require more, if the gut allows it. If the gut only takes 0.80 L/h, honest finish deficit is 2.80 L, 3.5% of 80 kg before any event urine. A 4% loss at 30 °C has cut cycling time-trial work by about 12%.⁹ Do not force 1.14 L/h on race day and call the vomit a hydration win. Gut training can raise the ceiling. It does not create a universal 1.0 to 1.2 L/h law. Gastric emptying varies from athlete to athlete, and hard efforts can slow it.¹⁴⁸

If 0.80 L/h is the honest ceiling, drink 0.80 and treat the rest after. Sodium at 50 mmol/L on that 1.60 L/h sweat is 1,840 mg/h lost. A drink at the top of ACSM's band, 690 mg/L, returns about 550 mg/h at 0.80 L/h. The gap is real. Closing it with a forced extra liter is how hyponatremia gets a head start.

A 2% deficit you cannot drink through is the case for glycerol hyperhydration, rehearsed in training, never debuted at the gun. If you do not know your sweat rate or cannot follow a set plan, drink to thirst.⁷ A tested drink rate belongs on hot, long, high-sweat days with restricted access, and it still carries a no-gain cap.

06Replace the rest after the session

Whatever you did not drink is a recovery volume, with sodium. Use the sweat-balance estimate of fluid deficit, including event urine and excluding known food or fuel mass, rather than raw post-event body mass. When nonfluid mass change is negligible, target 1.25 to 1.5 L of fluid for every kilogram of estimated fluid deficit.⁸ Shirreffs 1996 dehydrated twelve people by about 2%. Replacing 50% of the loss never restored balance, and 100% still left them short. One hundred fifty percent worked when the drink held 61 mmol/L sodium and leaked when it held 23 mmol/L.¹⁵

Use a salted meal and a sodium-containing drink. A chugged gallon of tap water is the leaky version. If another hard session sits inside 24 hours, this is the whole job. If you have two easy days, food and thirst will cover more of it. Weigh the next morning. The post-session plan failed if you are still well below the pre-session mass with another heat day coming.

07Who should skip the test or run it with a clinician

Skip this protocol, or run it only with a clinician, if you have kidney disease, heart failure, SIADH, a hyponatremia history, diabetes with unstable fluid handling, pregnancy, or a prescription diuretic. Diuretics make urine mass a lousy sweat proxy. Heart failure and SIADH make extra fluid a medical problem.

Nude or repeated weigh-ins can harm people with an eating-disorder history. Same-kit clothed weights still put a number on the scale twice a session. If the scale itself is the harm, do not run the test.

08Why weight gain during the event overrules the plan

Hew-Butler 2015 and the Wilderness Medical Society 2019 update both treat overdrinking of hypotonic fluid as the main preventable driver of symptomatic exercise-associated hyponatremia. Almost all of those cases gained or held weight.¹⁶¹⁷ Absence of gain does not clear you in an ultra. A rising scale is still a stop.

If a checkpoint or a portable scale is up, stop drinking. The L/h you calculated, the bottles you packed, and the sodium you mixed no longer apply. Collapse with confusion, headache, vomiting, or visible swelling is a call to EMS. Do not treat it as underdrinking. The plan exists to keep you below 2%. The scale decides when the plan stops.