How to Test Concrete Moisture
The test measures the building as much as the slab, and it takes days rather than minutes. A reading taken in a cold, shut up building describes a building that will not exist once the floor is in service.
The reading describes the building, not just the slab
Concrete exchanges moisture with the air above it, continuously, in both directions. That single fact is why a moisture test has a setup procedure at all, and why the procedure is mostly about the room rather than about the floor.
A slab in an unheated shell, with the doors open and the HVAC off, is in balance with outdoor air. The same slab in a closed, heated, occupied building settles somewhere else entirely. If you test the first one and coat the second, you have measured a building that is about to stop existing. Both ASTM F2170 (in situ relative humidity) and ASTM F1869 (calcium chloride) require the test area to sit at its intended service temperature and humidity for at least 48 hours before testing begins.
On a construction program that is a scheduling conversation rather than a technical one, and it is usually somebody else’s cost. Have it early, in writing. A reading taken outside those conditions is the first thing an expert will find if the floor fails, and it is not defensible.
The timeline is the part that changes your schedule
Neither test is an instrument you wave at a floor. Both are mostly waiting, and the waiting is not optional, so the practical shape of a moisture test is several days of elapsed time around a few minutes of work.
Day 0 · Before anything
- The building reaches its service temperature and humidity, and stays there.
- Both standards require 48 hours at service conditions before testing starts.
Day 2 · Prepare
- Drill and sleeve the probe holes, or grind the dish areas back to bare concrete.
- A calcium chloride dish on a sealed surface measures the sealer.
Days 2 to 5 · Wait
- Probes equilibrate in their holes. Dishes sit under their domes.
- Reading early reads low, which is the direction that gets a floor coated.
Day 5 · Read
- Take the readings, record the conditions, and keep the paperwork.
- The number has to be defensible months later, not just today.
The day numbers are the shortest honest version, assuming the building is already conditioned on day zero and that the probe equilibration period is the 24 hours the current revision of ASTM F2170 specifies. Older revisions asked for 72 hours and some kit makers still specify longer, so follow whichever period is longest. A calcium chloride dish sits for 60 to 72 hours regardless.
Where the hole goes, and why the depth is not arbitrary
A drying slab is not uniformly wet. It is driest at the surface, where moisture has been leaving, and wettest deeper down. So the depth you measure at decides the answer you get, and measuring near the top of a slab tells you about the part that has already dried.
ASTM F2170 sets the depth as a fraction of slab thickness rather than a fixed measurement, because it is chasing a point in the moisture gradient rather than a distance. For a slab drying from one side, which is a slab on grade or anything on a vapor retarder or a steel deck, the hole goes to 40% of the slab depth. For a slab that can dry from both faces, the hole goes to 20%.
The reasoning is worth carrying rather than memorizing. That depth approximates the moisture condition the whole slab will equalize to once you seal the top, which is exactly the condition your coating has to survive. A shallower hole measures the part of the slab that is about to stop being representative the moment a coating goes over it. This is also why the slab thickness needs to be known rather than assumed: get the thickness wrong and the depth is wrong, and the reading is answering a question nobody asked.
One test is not a survey
ASTM F2170 calls for three test locations for the first 1,000 ft² and one more for every additional 1,000 ft². ASTM F1869 scales the same way. That is a floor, not a target, and treating it as a target is how a floor gets coated over a wet corner nobody tested.
Moisture varies across a slab, and it does not vary randomly. It concentrates where the ground outside is wet, below a downpipe or a leaking gutter, along the base of walls, at construction joints, under racking and anything else that has stayed covered, and in whichever part of the building was finished last and has had the least time to dry. Place the required tests where those conditions are rather than on a tidy grid, and if you can only justify one extra location, put it in the worst corner of the building.
Two readings that disagree are information rather than a problem. They say the slab is not uniform, and the next question is how far the wet zone reaches. Three more test locations is a cheap way to answer that. A failed floor is not.
What quietly invalidates a reading
Every fault below produces a number. That is what makes them worth listing: none of them announces itself, and most of them read in the direction that gets a floor coated.
| What went wrong | What it does to the number |
|---|---|
| Testing before the building is conditioned | Measures an equilibrium the finished building will never sit at. Usually reads high in a damp shell and low in a dry windy one. |
| Reading a probe before it has equilibrated | Reads low, because the air in the hole has not yet come into balance with the concrete around it. |
| A dish on a sealed, coated or power troweled surface | Measures whatever is on top of the slab. Reads low, and reads differently in two places on one floor. |
| Dust left in the hole, or a sleeve that is not sealed | The hole exchanges air with the room, so the probe reports something closer to the ambient humidity than to the slab. |
| A kit past its date, or a scale that has not been checked | Calcium chloride is hygroscopic and does not keep. A drifted scale turns a weight difference of a few grams into the wrong answer. |
| Testing after the slab has been wetted | A wash down, rain through an open doorway, or a wet trade working overhead. The surface moisture is real but it is not the slab condition. |
| Testing only the open floor | The middle of a bay is the driest, best lit, most trafficked part of a building. It is the part least likely to fail. |
Down the side is the mistake. Across the top is what it does to the number you end up holding. Note how many of them read low, which is the direction that gets a floor coated.
What to write down
A moisture reading is evidence, and its whole value is that it can be produced later. If a floor blisters in eighteen months, the question will not be whether you tested. It will be whether the test you ran was valid, and that is answered by the record rather than by memory.
- The date and time of the reading, and the date conditioning started
- Where each test was, marked on a floor plan rather than described in words
- Slab thickness and hole depth for each probe location, since the depth is derived from the thickness
- Air temperature and humidity in the space, which is the evidence that service conditions were met
- How long each probe equilibrated, or how long each dish was down
- Kit, lot and calibration details, including the expiry on a calcium chloride kit
- Every reading, including the ones that passed, because a single quoted number with no context is worth very little
Put the date and the result on the quote as well. It sets out what the price was based on, and it makes a re test a visible change rather than an argument.
What the number does next
Testing produces a figure. It does not produce a decision. What the figure means, and whether the slab is clear, borderline, needs a barrier, or needs the source found before anyone coats anything, is a separate question with its own limits.
Judge the reading you just took
The procedure above gets you a defensible number. This is what it is measured against: enter the reading and the site conditions, and the tool returns a verdict for each test plus an overall go or no go.
Nothing is stored. The reading is a measurement of one slab on one day.
Open the live calculator →If the reading calls for a barrier, the barrier sizing guide covers what the two tests measure, the four bands a reading falls into, and how much material a barrier takes. If you want the other moisture question, the one about water condensing onto the slab on the day you coat, that is the dew point rule, and a slab can pass one of those checks and fail the other on the same morning.
What this page deliberately does not contain
It does not compare the two tests against each other, and it does not publish the limits a reading is judged against. Both already exist on this site, in one place each, and repeating them here would give you two versions to keep in step.
- What each test measures and why they disagree sits on the barrier sizing guide, alongside the decision the comparison feeds
- The limits themselves, and the four bands a reading falls into, belong to the moisture calculator, which is what judges a reading
- No table from either ASTM standard is reproduced here. The methods are described in our own words and the standards are named so you can buy them. Where a procedural figure appears above, it is attributed to the standard it comes from
Frequently Asked Questions
Can I test the slab on the day I plan to coat?
No, and this is the single most useful thing to know before you quote. Both recognized tests are mostly waiting.
An in situ relative humidity test to ASTM F2170 needs the space held at its service temperature and humidity for at least 48 hours first, then the probe has to sit in the hole and equilibrate before the reading means anything. The current revision of the standard puts that at 24 hours, older revisions asked for 72, and some kit makers specify longer, so follow whichever period is longest. A calcium chloride test to ASTM F1869 has the same 48 hour conditioning requirement and then leaves the dish down for 60 to 72 hours.
Either way you are looking at the better part of a week between arriving on site and holding a number you can act on. A test booked for the morning of the pour is not a test, it is a formality, and it will either delay the job or get skipped.
Does the building really need to be at service conditions first?
Yes, and it is the requirement most often ignored, because it is the one that costs someone else money.
Concrete exchanges moisture with the air above it. A slab sitting in an unheated shell in winter, or in a building with the doors open and the HVAC off, reaches a different equilibrium than the same slab will reach once the building is closed up, heated and occupied. Test it in the first condition and you have measured a building that is about to stop existing.
Both ASTM F2170 and ASTM F1869 require the test area to be at its intended service temperature and humidity for at least 48 hours before testing starts. In practice that means the HVAC running and the building closed, which on a construction program is a scheduling conversation rather than a technical one. Have it early. A reading taken outside those conditions is not defensible if the floor later fails.
How many places do I need to test?
More than one, and the standards set a floor rather than leaving it to judgment. ASTM F2170 calls for three test locations for the first 1,000 ft² and one more for each additional 1,000 ft², and ASTM F1869 scales the same way.
That is a minimum, not a target. Moisture varies enormously across a slab, and the variation is not random: it concentrates where the ground is wet, where a downpipe discharges, along walls, at construction joints, under anything that has stayed covered, and in the parts of a building that were finished last. A grid of tests placed for tidiness will miss all of that.
So take the count from the standard, then place the tests where the moisture is likely to be rather than where the tape measure says. If you can only argue for one extra location, put it in the worst corner of the building.
Can I run a calcium chloride test on a slab that is sealed or coated?
No. The test measures moisture arriving at the surface, so anything sitting on that surface is measuring the sealer rather than the slab, and it will read low.
ASTM F1869 requires the test area to be bare, clean concrete. In practice that means grinding a patch back to bare slab and letting it condition before the dish goes down, which is why the test area needs planning rather than a spare twenty minutes. A cure and seal that has broken down in the middle of a bay and survived at the walls is the case that catches people: the same slab can read clear where the film has gone and low where it has not, for reasons that have nothing to do with moisture.
This is also a good reason to establish what is on the slab before you plan the testing, not after.
Two locations gave different answers. Which one counts?
The worst one, and it is not close.
A coating is a single system laid over the whole floor, and it will fail wherever the slab beats it. Averaging a wet corner against a dry bay produces a number that describes no part of the floor and protects none of it. If one location says a barrier is required and three say clear, the floor needs a barrier, or it needs the wet area understood well enough to justify treating it separately and recording why.
The useful response to a disagreement is usually more testing rather than a decision. Two readings far apart is information: it says the slab is not uniform, and the next question is how far the wet zone extends. That is cheaper to answer with three more test locations than with a failed floor.
How long is a moisture reading good for?
Long enough for the job you tested for, and no longer than the conditions stay the same.
A reading is a snapshot of a slab in a building at a moment. If the job slips by a season, if the building is opened back up, if the heating is turned off over a shutdown, or if water gets in through a roof or a doorway in the meantime, the number you are holding describes a situation that has moved. Slabs also keep drying, so an old reading can be pessimistic as easily as optimistic.
There is no universal expiry date, so the honest rule is procedural: if anything about the building or the schedule has changed materially since the test, re test before you coat, and put the date of the reading on the quote so that everybody can see what it was based on.