Chocolate making, professionally
Cocoa Bean Quality Checks: Moisture, Cut Test, Bean Count
On this page
Checking cocoa beans on arrival takes three tests a small maker can run without a lab: moisture, a cut test and a bean count. You need a scale, a knife, good light and either a moisture meter or an oven. The published limits for each come from ISO 2451 and from trade contracts, and they mark what can be traded rather than what makes good chocolate. These checks open the bean-to-bar process, which follows the beans from here to the finished bar.
Most figures below come from Cocoa Beans: Chocolate and Cocoa Industry Quality Requirements, written by a working group of the European cocoa and chocolate trade associations and the Federation of Cocoa Commerce (FCC). It’s an industry guide rather than independent research, and the most accessible text that sets the ISO standard and the trade’s contract terms side by side. The full guide is online.
Moisture: where the limits sit
The industry guide expects beans at about 7% moisture, and the limits around that figure depend on whose standard you buy against. ISO 2451, the international specification for cocoa beans, requires beans dried to no more than 7.5%, according to the guide’s summary of the ISO cocoa standards.
The guide sets 8% as an absolute maximum. Its section on moisture content explains why both ends matter: above 8% you lose edible material and risk mold and bacterial growth, and below 6.5% the shell gets so brittle that beans break up.
A draft Caribbean regional standard, CROSQ DCRS 83, based on ISO 2451, narrows the range to 6 to 7.5%. The sources don’t settle on one target, so write down which one you buy against.
How to measure it
The CROSQ draft gives the reference method: grind the beans, dry them for 16 hours in a ventilated oven at 103 °C ± 2 °C and calculate the loss in mass. Most moisture meters work on capacitance. Whichever you use, the draft says the meter has to be calibrated for cocoa beans in the moisture range you’re testing, and the beans have to be at ambient temperature.
The cut test
The cut test is the trade’s quick look inside the bean. Under ISO 2451, as the guide’s Appendix A describes it, 300 beans are cut lengthwise through the middle, both halves are examined in daylight or equivalent light, and each type of defect is counted as a percentage of the 300. ISO defines nine defect categories, but only five feed into its grades: moldy, slaty, insect-damaged, germinated and flat.
The FDA method is different: 100 beans broken open, and only three categories (moldy, infested, or both). If you cut fewer beans than either method, your percentages can’t be compared with either.
Reading the cut surface
The trade definitions in Appendix A and the CROSQ draft tell you what you’re looking at. A slaty bean shows a slate color on at least half the cut surface and usually means the bean was badly under-fermented or not fermented at all. A violet or purple bean usually means some under-fermentation.
A bean counts as moldy when mold is visible to the naked eye inside it. Don’t confuse that with white spots, which the definition describes as a concentration of theobromine or cocoa fat.
The test has a clear limit. The guide’s introduction to trade quality standards says these checks cannot measure or ensure good flavor, although the cut test can detect gross flavor defects. Tasting a sample is still on you.
Cut-test limits: ISO, FCC and US contracts
ISO, the FCC and the US trade each set limits for the cut test, and they don’t agree:
| Framework | What it is for | Moldy | Slaty | Insect-damaged, germinated, flat |
|---|---|---|---|---|
| ISO 2451:2017, Grade I | Grading fermented beans in producing countries | 3% | 3% | 3% combined |
| ISO 2451:2017, Grade II | Same | 4% | 8% | 6% combined |
| FCC contract, “Good fermented” | Trade contracts | Counted as defective | 5% | 5% “defective” (moldy plus insect-damaged) |
| FCC contract, “Fair fermented” | Trade contracts | Counted as defective | 10% | 10% “defective” |
| CMAA contract (US), referring to the FDA standard | Trade contracts | 4% | Not listed | 4% infested or damaged; moldy and infested together no more than 6% |
Sources: ISO figures as reported in the guide’s ISO summary. FCC and CMAA terms as reported in the guide and its Appendix A.
The guide attaches its own warnings to these numbers. The frameworks don’t line up: beans with 3% insect damage and 3% mold can meet ISO Grade I and still exceed the FCC limit for Good Fermented.
More important, none of these is a quality bar for chocolate. The contract limits “do not imply acceptability for chocolate manufacture”, in the guide’s words. They are the levels at which allowances become due in arbitration. Manufacturers want beans “well within” Grade I, and the guide warns that cocoa very near the Grade I limits may hold as many as 8 or 9% defective beans and is unsuitable for good quality chocolate.
Bean count and size
Bean count is the number of whole beans per 100 g. ISO calls 100 or fewer “large” beans, 101 to 120 “medium” and more than 120 “small”, and the FCC splits the same range into four classes, according to Appendix A.
Size matters for yield. The guide’s page on bean size uniformity says a bean should weigh at least 1.0 g, because smaller beans carry more shell and less nib. It matters for processing too: by the guide’s rule of thumb, no more than 12% of beans should fall outside plus or minus one third of the average weight, because very mixed sizes are hard to clean.
Size spread follows the beans into the roaster. The guide’s roasting protocol only skips adjustments for size when beans run between 70 and 130 per 100 g.
Lab checks worth sending out
If you send samples to a lab, the guide gives two more reference points. Shell normally makes up 11 to 12% of the weight of main-crop West African beans, the norm other cocoas are judged against. And the fat of well-fermented, well-dried and properly stored beans generally has a free fatty acid content below 1%, and certainly below 1.3%. The EU legal limit for cocoa butter is 1.75%.
Sorting and cleaning
Sorting takes out what the cut test and your eyes flag: bean clusters, flat and broken beans, and anything that isn’t cocoa. Broken beans are worth removing early, because the guide notes that more broken beans and fragments can significantly raise the free fatty acid content of the fat.
The FCC sets thresholds for foreign matter (above 0.75% of the sample’s weight), flat beans and sievings (each above 1.5%) and cocoa-related matter (above 3.5%). Appendix A presents these as the points at which arbitrators may award a price allowance. They aren’t quality targets for a maker.
Cleaning is also where hygiene zoning starts. Raw beans carry an intrinsic risk of microbial contamination, and the guide calls for an effective hygiene barrier between incoming raw materials and finished goods. Its sample protocol for breaking and winnowing asks for precautions against cross-contamination between raw and roasted beans as part of an active HACCP program. In a small room, that comes down to separate containers, separate tools and a route that never brings roasted nibs back past raw sacks. The rest of the plan, from HACCP to the roast, is in chocolate food safety for small makers.
What to record for each lot
For every lot, write down the moisture reading and how you took it, the cut-test percentages by defect with the number of beans cut, the bean count per 100 g, and which framework you buy against. The moisture and bean count also go into the roast log, and the whole record is what you compare against when the next lot from the same supplier arrives.
Sources and open questions
ISO 2451 was read only through the industry guide and the CROSQ draft. The standard itself is paid, so which moisture method it specifies remains an open question.