The best hygrometers, ranked on published accuracy

Nearly every recommendation on this site resolves to the same instruction. Work out whether to run a dehumidifier — take a reading. Decide whether your house wants moisture added or removed — take a reading. Diagnose a window that streams every morning — take a reading, in that room, at the hour the problem is worst.

Thirteen pages on this site tell you to do that and none of them, until now, told you what to do it with. This is that page.

1. Published humidity accuracy, in ± % RH. The whole decision. A meter is an instrument, and an instrument with no stated tolerance is a decoration. Where a maker publishes both a typical and a maximum figure we use the maximum, because that is the one that decides whether a reading can be trusted.
2. The conditions attached to that figure. An accuracy number means nothing without the humidity range and temperature it applies at. "±2% from 10 to 90% RH" and a bare "±2%" are not the same claim, and we rank the first above the second even though they print the same number.
3. Whether it can be corrected. Humidity sensors drift. A manual calibration option is the difference between a meter that ages gracefully and one you eventually throw away.
4. Whether it records. A humidity problem is a pattern over nights and weeks, not a number at a moment. High/low memory is the minimum; real logging is better.
5. Where it can physically go. A probe or a remote sensor reaches the crawl space, the wall void and the cold cupboard that a wall gauge cannot.
6. What is missing. Battery life, sensor part numbers and IP ratings go unpublished across most of this category. We show those gaps rather than filling them.

We have tested none of these and own no reference hygrometer to test them against. Every figure here is the manufacturer’s own, quoted with the conditions they attach to it, and the arithmetic in the next section is marked as ours. How we pick.

The EPA publishes the target this whole site works to: keep indoor relative humidity “below 60 percent (ideally between 30 and 50 percent)”. That is a 10-point ideal band with a hard ceiling 10 points above it, and it is the reason accuracy is not an audiophile concern here.

If your meter can be out by a percentage points in either direction, then to know the true figure is under a limit, the display has to read the limit minus a. That is the whole calculation, and it turns an abstract tolerance into a number you can act on.

Read the bottom row again. With a meter whose published maximum error is ±7.5%, a display showing 55% is consistent with a true humidity anywhere from 47.5% to 62.5% — comfortably inside the ideal band, or over the EPA ceiling entirely. It cannot answer the question.

And read the top rows for the opposite reason: the gap between ±1.5% and ±2% moves the threshold by half a point. Beyond a certain tightness this stops mattering, which is why the cheapest meter here is the top pick rather than the dearest.

Every pick on this page publishes a ±2% humidity figure somewhere. They are not the same claim, and the differences are entirely in the conditions each maker attaches:

• SensorPush HTP.xw: ±2% typical with a separate ±3% maximum, stated across 0 to 100% RH, at 77 F — plus a tighter ±1.5%/±3% pair for the 0 to 80% range. Four numbers, two ranges, a temperature.
• AcuRite 01080M: ±2% from 10 to 90%, and ±4% outside it. One number, one range, and an explicit statement of where it degrades.
• Inkbird IBS-TH1 Plus: ±2%RH. No range, no temperature, no typical-versus-maximum distinction.

On the headline number those look identical. In information content they are not close, and the ranking on this page follows the disclosure rather than the digit. A maker who publishes a worse number with its conditions is telling you more than one who publishes a better number without them.

This is the same pattern this site keeps finding in cooling equipment, where Vornado publishes a throw distance and no wattage, and most portable fans publish no airflow at all. Which brands publish, measured across the fan category.

Nearly every modern dehumidifier shows a humidity percentage and lets you set a target against it. It is natural to treat that as the reading, and to conclude you do not need a separate meter at all.

The federal certification data says otherwise by omission. The ENERGY STAR Certified Dehumidifiers dataset records, for all 559 certified models: capacity on DOE Appendix X1, integrated energy factor, annual kWh, refrigerant type and global warming potential, refrigerant charge, and whether the unit is connected. It contains no humidity-accuracy field, and no acoustic field either.

So the percentage on the panel has no published tolerance behind it, and no maker we have looked at states one. It is good enough to cycle the compressor against; it is not good enough to tell you whether the room is inside a 10-point band. There is also a placement problem: the humidistat sits inside the machine, reading air it has just dried, which is not the air in the far corner of the room.

1. Above 60%: outside the EPA band, with a mold risk attached. Dehumidify, and look for a source — a dehumidifier run against active water ingress is a permanent bill with no end state. Dehumidifiers on certified efficiency.
2. 50 to 60%: inside the published maximum, above the ideal. Worth a dehumidifier on a humidistat rather than a continuous run. What to set the dial to.
3. 30 to 50%: the target band. If something is still wet — a window, a cold corner — the problem is a surface temperature rather than the air, and no machine on this site fixes it. Why glass goes wet in a dry house.
4. Below 30%: below the band. That is humidifier territory, which this site does not stock and will not pretend to review. Which machine you actually need.

One reading is also not a diagnosis. Humidity swings overnight as surfaces cool, and the figure that matters for mold and for condensation is the peak rather than the average — which is the practical case for high/low memory on a cheap meter, or logging on a dearer one.

• Away from the dehumidifier’s outlet. Standing a meter in the dried airstream reads the machine, not the room.
• Off external walls and away from windows. Those surfaces are the cold spots, and air against them is not representative of the room — though if the cold spot is what you are diagnosing, that is exactly where the probe should go.
• At roughly the height that matters. Moisture stratifies. A reading at floor level in a basement and one at head height can differ meaningfully.
• Not in direct sun and not above a radiator. Both raise local temperature, which drops relative humidity and flatters the room.
• Give it time. These sensors take minutes to equilibrate after being moved, and far longer after coming in from a cold car.