Making Pasta Filata
Pasta filata means spun paste, and it is the only family in the section where the defining step happens after the cheese is otherwise finished. You make an ordinary curd, and then you decide whether it becomes mozzarella or a bucket of ruined porridge.
The family page covers who belongs. This is how the stretch works and how to hit it.
The one thing: the stretch is a pH event, not a temperature one
Hot water doesn’t make curd stretch. Hot water makes curd stretch if the pH is right, and does nothing useful otherwise.
The mechanism is the one from melting, run deliberately. As the curd acidifies, calcium leaves the casein network. At high pH there is too much calcium still stapling the network together — heat it and it stays short and tears. At low pH the calcium is gone and there’s no network left — heat it and it collapses into curds in the water.
In between, for a while, the protein has just enough structure to be pulled and just enough freedom to align. That’s the window:
| Curd pH | What happens in the hot water |
|---|---|
| Above 5.6 | Tough and short. Tears rather than pulls. Wait. |
| 5.4 – 5.0 | It stretches. This is the window. |
| 5.2 | The middle of it — where I’d aim |
| Below 4.9 | Crumbles and dissolves. Too late; nothing brings it back. |
⚠ And the window is narrow in time as well as in pH. A curd that is fermenting steadily can cross from 5.4 to 4.9 in well under an hour. This family has a reputation for being temperamental, and that reputation is entirely about a cook who went to do something else for forty minutes.
🔴 Which is why two mozzarellas exist, and they are not the same cheese
Direct acidification — citric acid stirred into cold milk before the rennet — puts you at the right pH by arithmetic. You measure the acid, you land at the number, and the stretch is predictable enough to do on a weeknight. That’s the “30-minute mozzarella” of every beginner’s book, and it is genuinely reliable.
Cultured mozzarella gets to the same pH by fermentation, over hours, and tastes considerably better for it — the culture is producing flavour compounds the citric acid cannot. It is also the one that can overshoot while your back is turned.
So the trade is real and it isn’t skill-versus-shortcut: one is measured and safe, the other is grown and better. My own first cheese was the direct-acid version, and the cultured one is the obvious next thing on my bench rather than something I’ve already done.
The make, step by step
1. Ripen. A gallon of milk to 90–95 °F / 32–35 °C, thermophilic culture, hold 45–60 minutes. Thermophilic, not mesophilic — you’ll be at 105 °F shortly and a mesophilic culture dies there.
2. Rennet, and a firm set. Standard dose, diluted. Flocculation multiplier around 3 — you want a firmer, drier curd than a bloomy rind, because everything from here is about expelling whey.
3. Cut to ½ inch, rest five minutes, then stir gently while raising to 105–108 °F / 40–42 °C over about 30 minutes. Hold and stir until the curd is springy and the pieces don’t reknit when you squeeze a handful.
4. Drain, and let it mat. Pour off the whey, keep the curd warm — around 100–102 °F — and leave it as a slab. It will knit into a solid mass. This is where the acid develops, and it takes anywhere from 90 minutes to five hours depending on your culture and your room.
5. Test, don’t time. ⭐ This is the whole skill. Cut a golf-ball of curd, drop it in water at 175 °F / 80 °C for a minute, and pull. If it stretches into a smooth rope two or three feet long without breaking, the batch is ready. If it tears short, wait twenty minutes and test again. If it falls apart into shreds, you’ve missed it.
A pH meter turns this into a number rather than a judgement, and it’s the single piece of kit that does most for this family. The test is still worth running — the meter tells you where you are, the golf ball tells you what that pH is doing to your curd.
6. Mill and stretch. Cube the slab. Put it in a bowl and pour over water at 170–180 °F / 77–82 °C — or hot whey, which tastes better and is free. Work it with a spoon, then hands, folding the mass over itself until it is smooth and glossy.
Fold, do not tear. And stop as soon as it’s smooth.
7. Shape and cool. Pull a ball, tuck the seam underneath, and drop it into cold — ideally iced — water or brine to set the shape. A quick brine (about 5 minutes for a small ball in saturated brine) salts it; salting in the stretch water is the other option and gives a more even result.
What “right” looks like
The stretch water should stay clear. ⚠ If it goes milky white, you are losing fat, and that is the family’s characteristic waste. Water too hot, or worked too long, or both.
The finished ball should be glossy and slightly springy, and should weep a little whey when cut. A dry, matte, dense ball has been overworked.
Cut it and the interior should show fibres, faintly, running one way. Those aligned protein fibres are literally what “string cheese” is.
The members, and what changes
Fresh mozzarella (fior di latte). The base make above, eaten within days. Mozzarella di bufala is the same technique on water-buffalo milk, which is a different milk — far higher fat and protein, and the reason the real thing tastes as it does.
Low-moisture mozzarella. Drier curd, more thorough cook, salted more, then aged briefly. This is the pizza cheese, and it is the correct pizza cheese for the reason on the melting page: stretch needs intact protein, so you want young, and you want low moisture so it browns rather than floods.
Burrata. A mozzarella pouch filled with stracciatella — shredded stretched curd mixed with cream — and closed at the top. It’s an assembly, not a separate make, and it was invented to use up the offcuts.
Provolone. Stretched, then aged, and traditionally made with a rennet paste carrying lipase. The lipase does lipolysis on the milk fat and produces free fatty acids: that’s the piccante bite. Dolce is the same cheese, younger and with less of it. If you want the bite at home you add lipase powder at the culture stage — a little goes a long way, and too much reads as rancid rather than sharp.
Scamorza. Mozzarella dried for a week or two, often smoked — see smoked and flavoured.
Oaxaca (quesillo). Stretched into a long ribbon and wound into a ball. Same curd, different hands.
String cheese. The fibres pulled in one direction and kept there. The novelty is real food science: you are aligning a protein network on purpose.
Kashkaval, caciocavallo. The eastern-Mediterranean and southern-Italian aged members — same stretch, months of aging, and much closer in eating to a hard cheese than to mozzarella.
What goes wrong
It won’t stretch, it tears. pH too high. This is recoverable — put the curd back somewhere warm and test again in twenty minutes. Nearly everyone’s first failure is impatience, not error.
It falls apart in the water. pH too low. Not recoverable as mozzarella; drain it, salt it, and eat it as a fresh cheese, which is a perfectly good outcome that nobody tells you about.
Tough, dry, squeaky. Overworked, or the water was near boiling. Once the fat is in the water it isn’t coming back.
It weeps and goes flat overnight. Under-drained curd, or not cooked firm enough at step 3.
A cultured batch that never gets there. Culture dead or dosed short, or the room too cold for the mat to hold 100 °F. Thermophilic cultures do very little at 75 °F.
The formulas
This page’s recipes are written:
- Cheese, Mozzarella (Cultured) — the make described above, with the golf-ball test as its own step
- Cheese, Burrata — an assembly rather than a cheese, and one invented to use up the offcuts
- Cheese, Provolone — the only member of this family that sees a cave, and the one where the bite is an enzyme you add on day one, not the age
- Cheese, Scamorza — the same curd, hung and dried, and usually smoked
- Cheese, Oaxaca (Quesillo) — the ribbon, and an origin story I have deliberately hedged
- Cheese, String — where the repeated one-directional pull is the whole recipe
- Cheese, Kashkaval — a two-day make, and the one that shows this family reaches a lot further east than Italy
⚠ And be exact about the one that was already here. Mozzarella Cheese (Traditional & Quick No-Rennet) is the acid-set make — it is the cheese I made first, it is very good, and it lives with the fresh cheeses. It is not the cultured cheese described above.
🔴 All seven of the new ones are marked UNTESTED, and that is not modesty. The only cheese in this family I have actually made is the direct-acid mozzarella. Each recipe says so in its first line.
Still owed: low-moisture mozzarella for pizza, and caciocavallo.
But two auxiliaries every make on this page needs are written: Cheese Culture, Mesophilic Mother and Cheese Brine, Saturated — and the brine one is the more important of the two, because a plain salt brine takes a good cheese apart in about a week.
Next: soft-ripened and bloomy rinds — running P. candidum without ending up at ammonia.
Comments (0)