How to measure hydrogen water at home — and why most ppm comparisons are wrong

What ppm, mg/L and mM mean

For dissolved hydrogen in water, 1 ppm = 1 mg/L = roughly 0.5 mM. They are the same quantity in three units.

Parts per million is a mass ratio: one milligram of hydrogen gas dissolved in one litre (one kilogram) of water is one part per million by mass. Scientific papers usually report mg/L or millimolar (mM); hydrogen has a molar mass of about 2 g/mol, so 1 mg/L is about 0.5 mM and 1.6 mg/L is about 0.8 mM. Marketing copy prefers ppm because the numbers look bigger. When you see "12 ppm per tablet", read it as "12 mg of hydrogen per litre" and then ask which litre, at what temperature, and measured when.

Why 1.6 ppm is a ceiling, and how tablets exceed it

At normal air pressure and room temperature, water holds at most about 1.6 mg/L of hydrogen at equilibrium. Anything above that is supersaturation: a temporary state that off-gasses within minutes once the container is open.

Hydrogen is a very poorly soluble gas (Henry's law constant about 0.78 mM/atm at 25 °C). A hydrogen water machine, which bubbles or electrolyses hydrogen into open water, cannot push past that ceiling and typically delivers 0.5 to 1.2 ppm. An effervescent magnesium tablet is different: it generates hydrogen inside the water, faster than the gas can escape, so for a few minutes the water carries more hydrogen than it can hold. In a sealed bottle the pressure rises and the supersaturation lasts longer; in an open glass it decays quickly. This is why tablet brands quote 6 to 12 ppm while machines quote 1 to 1.5 ppm, and why the two numbers only mean something alongside the container, volume, temperature and time of measurement. Our Science page covers what a PPM claim must state to be comparable.

Method 1: titration drops (methylene blue)

Titration kits are the cheapest reliable home method (about $40 for 100+ tests) and are accurate to roughly ±0.1 ppm if you work fast.

The reagent is methylene blue with a platinum catalyst. Dissolved hydrogen reduces methylene blue to its colourless form; once the hydrogen is used up, the next drop stays blue. The procedure:

1. Draw the sample immediately. Use the kit's 6 mL vial or syringe and take the water from the middle of the glass or bottle, within one minute of the tablet finishing.
2. Add one drop, cap, invert twice. If the blue disappears, hydrogen is present. Do not shake vigorously; agitation drives hydrogen out.
3. Keep counting. Add drops one at a time until a drop stays faintly blue for 10 seconds or more.
4. Multiply. Drops that cleared × 0.1 = mg/L (ppm). Twelve clear drops = 1.2 ppm. The drop that stayed blue does not count.
5. Record the conditions. Water volume, temperature, open or sealed, and seconds between tablet finishing and the first drop.

Limits: the reagent degrades in heat and sunlight, so store it cool and dark; the end point is a judgement call in the last drop; and the test consumes the sample, so you cannot re-measure the same water.

Method 2: dissolved-hydrogen meters

A true dissolved-hydrogen meter (such as the Trustlex ENH-1000 or a polarographic H₂ probe) reads continuously in ppb or ppm, costs $250 to $2,000, and needs frequent calibration.

These meters use an electrode that is selectively permeable to hydrogen; the current produced is proportional to dissolved H₂. The advantage is that you can watch the number climb and fall in real time, which is the only way to see how fast an open glass loses hydrogen. The disadvantages are drift, sensitivity to flow and temperature, and a long settling time (30 to 90 seconds) during which hydrogen is already escaping. Consumer meters also top out around 2 to 3 ppm, so they cannot read a supersaturated tablet sample directly; the sample must be diluted with hydrogen-free water and the result scaled, which introduces its own error. Laboratories use gas chromatography with a thermal-conductivity detector (GC-TCD) on sealed headspace vials, which is the reference method and is what we will use for our published independent test.

What ORP meters really measure

An ORP (oxidation-reduction potential) meter does not measure hydrogen concentration. A reading of −500 mV does not mean "more hydrogen" than −400 mV.

ORP responds to every redox-active species in the water: dissolved hydrogen, but also chlorine, dissolved oxygen, pH and minerals. The relationship between ORP and hydrogen is logarithmic, so going from 0.1 ppm to 1 ppm might move the reading from −300 mV to −400 mV, and 1 ppm to 10 ppm only another 60 mV. Negative ORP confirms that hydrogen is present; it cannot tell you how much. Any brand that reports its hydrogen output only as an ORP value has not measured hydrogen.

Seven things that change your reading

• Volume. The same tablet in 8 oz reads higher than in 16 oz; the hydrogen is spread through half the water. "Per tablet" ppm claims are always volume-dependent.
• Temperature. Cold water holds more gas, but a magnesium tablet reacts more slowly in cold water. Room temperature is the usual compromise, and the one Aloha H2's directions specify.
• Sealed or open. A capped bottle keeps supersaturated hydrogen for several minutes; an open glass can lose a third of it in two minutes.
• Time. Measure at 30 seconds and at 5 minutes and you will get two different numbers from the same glass. Always report the delay.
• Pouring. Transferring water to a second container strips hydrogen. Measure in the dissolving vessel.
• Surface area. A wide bowl off-gasses faster than a narrow bottle of the same volume.
• Agitation. Stirring or shaking after the fizz drives hydrogen out, not in.

Five mistakes in brand comparisons

1. Comparing a sealed-bottle number with an open-glass number. Not the same test.
2. Comparing different volumes. A 12 ppm claim in 250 mL and an 8 ppm claim in 500 mL describe the same total hydrogen.
3. Using ORP as a proxy. See above.
4. Measuring after pouring into a measuring cup. The pour alone can halve the reading.
5. Testing once. Tablet-to-tablet variation exists. Three tablets from the same pack, same protocol, averaged, is the minimum.

A fair test protocol you can copy

If you want numbers that can be compared across brands, hold everything constant except the tablet:

That protocol will give lower numbers than any brand's marketing figure, for every brand, because it uses a realistic volume and an opened container. That is the point: it measures what reaches the drinker.

Frequently asked questions

Last reviewed October 2026. Sources: Henry's law solubility data for H₂ in water (NIST Chemistry WebBook); Ohta S, Pharmacology & Therapeutics 2014 (dissolved-hydrogen measurement methods); H2Blue reagent documentation.