Sweat sodium varies several-fold between athletes and can be estimated reasonably well, but no trial shows that fuelling guided by a sweat test improves performance or prevents hyponatremia
Interpretation
Variation is well documented: in 506 athletes predicted whole-body sweat sodium ranged 18.2-70.8 mmol/L and sweat rate 0.26-5.73 L/h (Baker 2016); endurance athletes among 1,303 lost 51.7 mmol sodium per hour on average (Barnes 2019), though both datasets are commercial and mix test conditions. Under fixed conditions, sweat sodium was fairly stable and one 40-min sample predicted a 5 h ride average in 8 trained men (Goulet 2026). Forearm and thigh patches predicted whole-body sodium (r² 0.73 and 0.77) but not sweat rate (Baker 2018); body mass change is the best sweat-rate method (Baker 2017). Consumer devices read about 10 mmol/L low (53 vs 63 mmol/L; Bandiera 2026) and varied 23-24% between days (Zhao 2026). On outcomes, modelling suggests targeted replacement matters only with salty sweat (40 mmol/L or more) and more than 80% fluid replacement in very long events (McCubbin 2023); full personalised replacement in 9 runners over 5 h at 30°C changed plasma sodium slightly but not fluid balance, heart rate, core temperature or effort (McCubbin and Costa 2024); sodium tablets made no difference in 413 Ironman finishers (Hew-Butler 2006) and sodium intake was unrelated to ultramarathon performance in 266 runners (Lipman 2021). Hyponatremia (below 135 mmol/L) is driven mainly by overdrinking (Bennett 2020). No trial of sweat-test-guided fuelling in triathletes.
Practical implication
Measure sweat rate with before-and-after weighing, treat consumer sweat sodium readings as approximate, and use a result to choose roughly how salty race drinks and foods are, not to set a drinking target. Testing is most plausibly useful for long-course athletes racing in heat who drink heavily; for most athletes sodium can be chosen by taste and tolerance. Benefits for triathlon performance are inference.
Sources
12 sources, independently verified.
J Sports Sci 34(4):358-368. Retrospective analysis of forearm-patch sweat tests from one sports science institute. Predicted whole-body sweat sodium 35.9 ± 10.4 mmol/L (range 18.2-70.8); whole-body sweat rate 1.21 L/h (range 0.26-5.73). Wide between-athlete variation, partly reflecting varied test conditions. Data from an institute linked to a sports drink company.
J Sports Sci 37(20):2356-2366. Update of the same institute's dataset. Endurance athletes averaged sweat rate 1.28 L/h and sweat sodium loss 51.7 mmol/h, higher than every sport except American football. Wide individual ranges; commercial dataset.
Sports Med 47(Suppl 1):111-128. Review of sweat testing methods. Body mass change (corrected for intake and output) is the simplest and most accurate field method for whole-body sweat rate. Sweat sodium results depend on collection site, timing and duration, skin cleaning, sample storage and analysis. Within-person sweat sodium varies with intensity, environment, heat acclimation, fitness, body size, sex, age, diet and hydration.
Nutrients 18(6):989. Two 5 h cycling trials at 30°C with ad libitum or programmed drinking; forearm patch samples four times per ride. Whole-body sweat sodium varied by about 5.6-7.2 mmol/L within and between trials; a single 40-min sample predicted the ride average (R² = 0.87); drinking strategy made no difference. Small, male-only lab sample.
J Appl Physiol 124(5):1304-1318. 90 min cycling in the heat; patches at nine sites compared with whole-body wash-down. Regional patch sodium predicted whole-body sweat sodium (r² 0.55-0.77; forearm 0.73, thigh 0.77), but regional sweat rates did not predict whole-body sweat rate accurately enough to guide drinking.
Front Physiol 16:1717275 (published January 2026). About 74 min cycling in hot-dry (40°C) and hot-humid (30°C, 81% RH) conditions. A wearable sweat sensor and a handheld analyser both read sweat sodium of 53 mmol/L against 63 mmol/L by laboratory flame photometry. Wearable sweat-loss estimate 1.479 kg vs 1.432 kg by body mass; an online calculator underestimated (1.202 kg). Authors advise caution; devices not equivalent to laboratory analysis.
Appl Physiol Nutr Metab 51:1-9. Two matched 1 h sessions in warm, humid conditions; forearm patch sweat analysed by two portable analysers and a laboratory analyser. Strong association with the laboratory method (R² 0.867 and 0.852), but between-trial coefficient of variation of 23% and 24% for the two devices.
Eur J Sport Sci 23(6):992-1000. Modelling across sweat rates 0.5-2.5 L/h, sweat sodium 20-80 mmol/L and fluid replacement 10-90%. Sodium needs were driven by fluid replacement and sweat sodium. Replacement unnecessary in realistic soccer and elite marathon scenarios; in a 160 km ultramarathon targeted replacement was needed only with sweat sodium of 40 mmol/L or more and more than 80% fluid replacement. Otherwise 'season to taste'. Modelling, not an outcome trial.
Int J Sports Physiol Perform 19(2):105-115. Crossover: 5 h treadmill running at 30°C, ad libitum drinking, with capsules replacing 100% of measured sweat sodium losses or placebo. Plasma sodium rose 4 vs 1 mmol/L; no differences in fluid balance, heart rate, core temperature, thermal comfort or perceived exertion. In the 24 h after, the sodium trial drank more and excreted more sodium. Performance not measured.
BJSM 40(3):255-259. Randomised sodium (620mg), placebo (596mg starch), or no supplementation; no significant between-group differences in serum sodium, finishing time, or body-weight change — sodium tablets alone did not prevent or meaningfully affect hyponatremia risk in this trial.
Sports Med Open 7(1):13. Sodium supplement intake was not associated with pace, race time or finishing rank. Runners who were overhydrated performed worse than those who were dehydrated. Observational; sodium not tailored to sweat tests.
Wilderness and Environmental Medicine 31(1):50-62. Expert-panel guideline defining exercise-associated hyponatremia as blood sodium below 135 mmol/L during or up to 24 hours after prolonged activity. Its signs often overlap with heat exhaustion and exertional heat stroke, and failure to tell them apart is a recognised cause of worse outcomes; rapid recognition is essential in severe cases.