How much capping tells you depends on where you keep bees. In a dry climate it is a decent proxy — here on Colorado's Front Range I can pretty much assume a capped frame is under 18% moisture. In a humid climate the cap stops being evidence of anything — capping can close over honey that is still too wet, and pulling it on the assumption that capped means ripe is how a crop ferments in the jar. Uncapped honey needs checking everywhere, every time. My working limit is 18%, and the instrument that settles it costs less than one super of honey is worth — I think every beekeeper should own a refractometer.

Why capping and moisture come apart

Bees cap a cell when they have finished with it, and what "finished" means depends on the conditions they were working in. In a humid stretch, comb can be capped over honey that is wetter than you want, because the colony reached its own equilibrium rather than yours.

Two more ways the two come apart in practice:

Capped cells are sealed off from the air; uncapped cells are not. Open cells sit in equilibrium with the air in the super. Leave a partly-capped super on through a wet week and the uncapped portion gets wetter, while the capped portion stays where it was.
You extract the whole super at once. Even if nine frames are dry, the two wet ones go into the same bucket. Fermentation does not care that the average was fine.

The practical consequence: the thing you are measuring is not "is this frame capped" but "what is the wettest honey I am about to put in a bucket."

The number the grading standard actually sets

The USDA Agricultural Marketing Service standard for extracted honey (effective 23 May 1985) sets the analytical factor as a minimum percent soluble solids, determined with a refractometer at 20 °C:

U.S. Grade A — 81.4% soluble solids minimum
U.S. Grade B — 81.4% soluble solids minimum
U.S. Grade C — 80.0% soluble solids minimum

The standard's own conversion table puts 81.4% soluble solids at 18.6% moisture, and 80.0% at 20.0% moisture. So "18.6% for Grade A" is right, with two details worth knowing: Grade B carries the same analytical requirement — the grades differ on flavour, defects and clarity, not on moisture — and the standard is written the other way round from how beekeepers talk about it.

This is a grading standard, not a food-safety limit. It tells you what you may call Grade A. It does not tell you your honey will not ferment.

Why my limit is 18% and not 18.6%

I work to 18% rather than to the 18.6% grading line, and the gap is margin rather than theory: margin for the frames I did not sample, for a warm garage in August, and for a bucket that sits until Christmas. A super that reads exactly at the grading limit has none of that room.

Fermentation is not a switch that flips at a number anyway — it also depends on the yeast already in the honey and on how warm it is stored, which is why two buckets at the same reading do not always behave the same way. If you sell at volume, a lab that runs an osmophilic yeast count will tell you more than any moisture figure I could quote you.

What to do with a super that reads high

The cheapest answer is to put it back. If there is still a flow and the weather is dry, the bees will finish it themselves, and in a dry climate they usually do — which is why reading the super before the extractor is set up saves the most work.

If it has to come off, dry it before you extract: a stack of supers in a closed room with a dehumidifier and a fan moving air through the boxes will pull moisture out over a day or two. Airflow through the stack matters more than the size of the dehumidifier; a stack with no path through it just gets warm.

Whatever you do, do not let wet honey into the main crop. It is fine for feeding back or for mead, but blended into the rest it drags the whole batch toward the wet end.

Where you keep bees changes how much the cap is worth

Dry climate (what I run). On the Front Range, capped honey is reliable enough that I treat a fully capped frame as under my limit without testing it. That is not a claim that capping guarantees ripeness — it is a claim about the ambient air my bees are working in. Uncapped honey still gets tested, every time, before it goes near the extractor.

Humid climate. The same shortcut is dangerous, because the colony reaches its own equilibrium with much wetter air than mine and caps over honey that would not clear my limit. Keeping bees somewhere like the Southeast, I would test the capped frames too. I have not measured that myself — I keep bees in a dry climate — but the cap is not doing the work there that it does here, and treating it as though it were carries a real fermentation risk.

I extract uncapped frames that read under 18%

Cap state does not decide it for me; the reading does. Some keepers will not extract an uncapped cell at all, and there is a shake test — hold the frame horizontally, snap it, and see whether nectar flies out — that is genuinely useful at the very wet end. It just cannot tell comfortably dry from marginal, which is the band where the decision is actually hard. That is what the refractometer is for, and it is why I would rather own one than argue about capping.

Sources and further reading

U.S. Department of Agriculture, Agricultural Marketing Service — United States Standards for Grades of Extracted Honey (effective 23 May 1985). https://www.ams.usda.gov/grades-standards/extracted-honey-grades-and-standards — the 81.4% and 80.0% minimum soluble-solids requirements by grade, and the soluble-solids to moisture conversion table used above. Note this is a grading standard, not a food-safety limit.

Everything else is my practice at sideliner scale on Colorado's Front Range.

Capped Doesn't Mean Ripe: Reading Honey Moisture Before You Pull Supers