hDrop sweat sensor calibration: why the fluid loss numbers run high

The hDrop sweat sensor has appeared in three articles on this site. In each one, I flagged that the fluid loss figure looked too high. Today I ran a proper calibration test during a HYROX half sim, i.e., a controlled pre- and post-weigh-in with no fluid consumed. The result confirmed what I suspected.

Pre-workout nude weight 73 kg, post-workout 71.5 kg. Weight loss 1.50 kg, equal to 1.50 litres over approximately 100 minutes. hDrop measured 2.055 litres: a 37 per cent overestimate.

Three tests, same direction

1. Stockholm HYROX sim, June: body weight change roughly 2.5 kg, hDrop said 3.1 litres. Weigh-in check was not rigorous.
2. Dartmoor Classic: 108 km, 26°C, four hours. hDrop recorded 6.63 litres, I felt high by one to two bottles. Sweat rate did not drop during a 30-minute shaded stop. No controlled weigh-in.
3. Today: controlled conditions, zero fluid intake, clean weigh-in. Overestimate: 37 per cent.

Only today’s session had a proper weigh-in.

Why does the error happen?

Local-to-whole-body conversion. hDrop measures sweat at one site, the upper arm, and applies a fixed ratio to estimate total body fluid loss. That ratio shifts with intensity, temperature, clothing, acclimatisation, and individual sweat distribution. A ratio calibrated for moderate cycling under lab conditions will yield a different result during a high-intensity gym session.

Indoor pooling. hDrop’s electrodes need liquid sweat to function. Outdoors, sweat evaporates, and skin stays relatively dry between production cycles. In a low-airflow gym, sweat pools and keeps the electrodes saturated, so the algorithm reads a higher local rate than the body-mass change supports. Body-mass change is the best practical field estimate of whole-body sweat loss because it also captures evaporated fluid.

Sweat rate and direction of error. Precision Fuel and Hydration’s independent lab test found hDrop underestimated the sweat rate on a moderate cycling protocol. At higher rates in humid conditions, the error can flip. hDrop added an adaptive calibration feature in July 2025 for athletes losing more than 1.5 L/h. Today’s rate was around 0.9 L/h, below that threshold, so indoor pooling and the local-to-whole-body ratio are the more likely drivers here.

The correction factor

In my case, I divide the hDrop figure by 1.37. Applied illustratively to Dartmoor, 6.63 ÷ 1.37 gives approximately 4.8 litres: consistent with drinking 3.7 litres and finishing slightly dehydrated. The correction will vary with environment and intensity.

What the device gets right

Sodium likely held up better. At Dartmoor, concentration plateaued at 57 to 65 mmol/L and held there for the final two hours. Electrode-based ion concentration measurements are more direct than scaled whole-body fluid estimates.

Calibration

hDrop states no calibration is required. Three sessions say otherwise, at least for this athlete in these conditions (n=1).

For the full review, I will use the calibrated values. Further reading: Precision Fuel and Hydration’s independent hDrop test and hDrop’s science page.

Take Out

The measured sweat rate still exceeded my intuition of how much I sweat. A calibrated hDrop will certainly flag when I’m underhydrated by a whole bottle or more. Half a bottle out makes little practical difference, and I feel errors in measurement won’t exceed that sort of level.

The app workflow before each session is fiddly, but livable for periodic long, hot rides where hydration matters.

If you wear a biceps strap such as the Polar Verity Sense, hDrop clips on for dual use.

In summary: flawed but handy, calibration required.

FAQ

Does hDrop need to be calibrated?

hDrop says no. Three sessions here all showed overestimation, with the one controlled test returning 37 per cent. A nude weigh-in before and after a session with zero fluid intake gives a provisional personal divisor and makes the fluid-loss figure meaningful.

Why does hDrop overestimate fluid loss indoors?

The conductivity electrodes need liquid sweat. Outdoors, sweat evaporates and skin stays drier. In a low-airflow gym, sweat pools and keeps the electrodes saturated, likely causing the algorithm to read a higher local rate than actual whole-body loss warrants. Body-mass change captures evaporated fluid too, which is why it returns a lower total.

How do you calculate sweat loss from body weight?

Weigh naked before the session. Exercise with no fluid intake. Weigh naked immediately after. The kilogram difference approximates litres lost. If you did drink, add that volume to the weight loss figure.

Is the hDrop sodium data more reliable than the fluid loss data?

Based on these sessions, yes. The electrode measures ion concentration directly at the skin, a more direct problem than scaling a local sweat rate to whole-body fluid loss. Sodium is not subject to the same pooling and evaporation effects.

What is a normal sweat sodium concentration?

Most endurance athletes fall between 20 and 80 mmol/L, with the majority between 30 and 60 mmol/L. The 57 to 65 mmol/L recorded at Dartmoor sits at the higher end of moderate and points to a sodium-dense electrolyte strategy for long hot sessions.

Last Updated on 24 June 2026 by the5krunner