What Your Cheese Yield Number Actually Tells You

Ask how much cheese you get from a batch of milk and the usual answer is "about ten percent". That is a decent rule of thumb for one specific case and misleading for most others, because yield depends on three things that vary enormously: the fat in the milk, the casein in the milk, and how wet you leave the finished cheese.

The formula that handles all three has been in use since 1949.

Van Slyke, and what it says

The yield calculator uses the Van Slyke formula, the standard dairy-science estimate:

Y = [(0.93F + P − 0.1) × 1.09] ÷ (1 − M)

where F is the milk's fat percentage, P its casein percentage, M the target moisture fraction of the finished cheese, and Y the yield in kilograms of cheese per 100 kilograms of milk.

The pieces are readable once you know what they represent. The 0.93 is the share of fat actually retained in the curd rather than lost to the whey. The 0.1 subtracted from the casein accounts for casein that does not end up in the cheese. The 1.09 accounts for the salt and other solids the cheese picks up. And dividing by (1 − M) is the important one: it converts dry matter into finished weight, which is why moisture dominates the result.

Moisture is the biggest lever

Take ordinary whole cow milk at 3.6% fat and 3.2% protein and vary only the target moisture:

Finished cheese	Moisture	Yield per 100kg milk
Hard, aged	38%	10.10 kg
Semi-soft	50%	12.52 kg
Soft	55%	13.91 kg

Same milk, same everything, and the soft cheese produces 38% more finished weight than the hard one. Nothing about that is a gain in dry matter — it is water, retained rather than pressed and dried out.

This is worth internalising because it dissolves a common confusion. A recipe that yields more is not necessarily extracting more from the milk; it may simply be holding more water. The dry matter — the actual cheese — is nearly identical.

The milk matters too, and not equally

Now hold moisture at 50% and change the milk:

• Whole cow milk (3.6% fat, 3.2% protein): 12.52 kg
• Semi-skimmed (1.5% fat, 3.3% protein): 8.43 kg
• Sheep milk (7.5% fat, 5.8% protein): 22.59 kg

Skimming the fat costs about a third of the yield even though the protein barely moved — because fat is roughly as important as casein in the numerator, and there is a lot less of it in semi-skimmed milk.

Sheep milk nearly doubles whole cow milk. This is the arithmetic behind why sheep-milk cheeses are priced the way they are, and why a recipe written for cow milk produces a startling amount of cheese if you run it on sheep milk without adjusting anything.

The protein figure on the carton is not casein

One conversion catches people out. Recipes and dairy science want casein; milk cartons print total protein, which includes whey proteins that do not end up in the curd.

Casein is roughly 78% of total protein in cow's milk, and the calculator applies that share for you: enter the carton's 3.2% protein and it works from 2.50% casein. If you have an actual lab figure for casein, use it directly — but the 78% conversion is close enough that the difference is usually smaller than the variation between two cartons of the same milk.

What it is genuinely useful for

Buying the right amount of milk. Working backwards from a target is the most common use. If you want a kilo of hard cheese at 38% moisture from whole cow milk, you need roughly 10 kilograms of milk — the calculator gives 1.01 kg from 10 kg, which is about as close to a round number as this ever gets.

Diagnosing a make. If the formula predicts 10% and you got 6%, something went wrong and the number tells you to look for it. Common causes: curd lost through cloth, a set that never firmed properly, over-cutting the curd so fines washed out with the whey, or milk that was not what the carton claimed.

Comparing milks honestly. Farm milk that costs more per litre may be cheaper per kilo of cheese if its fat and protein are meaningfully higher. That is an arithmetic question and the calculator answers it.

Where the estimate stops

It is a back-of-envelope figure, and the sources of error are real:

• Vat losses are not modelled. Fines lost in the whey, curd stuck to equipment, and trimming all reduce the real figure below the prediction, and a small batch loses proportionally more than a large one.
• Finished moisture is a guess until you measure it. You are feeding the formula a target, not a measurement, and a cheese that ends up drier than intended yields less than predicted.
• Milk varies seasonally in both fat and protein, sometimes substantially, so a carton's printed averages are exactly that.
• Technique matters more than the formula admits. Cutting size, cooking temperature, stirring and pressing all move the moisture, which moves the yield.

Treat it as a planning figure that tells you roughly how much milk to buy and whether a finished batch was in the expected range. If your real yields sit consistently below prediction, the gap is your process, and it is a more useful thing to know than a single number ever is.

One thing yield does not tell you

A high yield is not a better cheese, and it is not a safer one. Moisture is the variable doing most of the work in this formula, and higher moisture also means a more hospitable environment for things you did not culture on purpose — which is why soft, high-moisture styles carry more caution than hard aged ones. Yield is an economic and diagnostic number. The safety questions are separate, and worth reading before a first make rather than after one.