How to Make Gelato at Home Without a Machine: Full Method
Table of contents
You need a freezer, a shallow metal tray, a fork and about six hours, of which roughly twenty minutes is work. What you do not need is a machine. What you cannot skip is the recipe, because a home freezer forgives nothing that a batch freezer would have hidden.
What Actually Makes Gelato Gelato
The equipment list is short and none of it is specialised: a digital scale that reads to one gram, a saucepan, a thermometer, a shallow metal tray or loaf pan, and a fork. A blender helps for one of the five methods below and is optional for the rest. The scale is the only item that is genuinely non negotiable, because a gelato recipe is a set of percentages and you cannot hit percentages with cups.
Three things separate gelato from everything else in the frozen aisle, and only one of them involves equipment.
Less air. Gelato is churned to roughly 25 to 35 percent overrun, against 50 to 100 percent for industrial ice cream. Less air means denser texture and more flavour per spoonful, and it is the single easiest specification to hit without a machine, because a fork whips in almost nothing.
Less fat, more milk solids. A gelato base runs about 6 to 8 percent fat where American ice cream runs 12 to 18. The body that fat is not providing has to come from somewhere, and it comes from MSNF, the non fat milk solids. This is the trade at the centre of gelato against ice cream, and it is why milk powder appears in every serious gelato recipe.
Warmer service. Gelato is served at about -12 degrees Celsius, ice cream nearer -18. Four degrees sounds trivial and is not: at -12 a correctly balanced gelato is soft, aromatic and scoopable, and at -18 the same tub is a brick. The serving temperature is a formulation decision, not a freezer setting, because it is the sugars in the recipe that decide how much of the water is still liquid at any given temperature.
That last point is the whole reason this page starts with numbers instead of a method.
The Eight Numbers a Gelato Recipe Has to Close
A machine mostly hides a bad recipe by freezing fast and whipping hard. Without one, the recipe has to be right. These are the eight figures, each in a sentence, each with the page that goes deeper.
| Number | Target | What it does |
|---|---|---|
| Fat | 6 to 8% | Carries flavour and slows melting, and too much makes the gelato heavy and greasy |
| MSNF | 10 to 11% | Milk protein and lactose that bind water and give body without fat |
| Sugars | 18 to 19% | Sweetness, solids and, crucially, the freezing point |
| Other solids | 2 to 4% | Maltodextrin, inulin or fibre that add body with no sweetness |
| Total solids | 38 to 42% | Everything that is not water, and the number that most directly predicts iciness |
| PAC | 270 to 290 | How far the sugars push the freezing point down, which sets the scoop at -12 |
| POD | 16 to 22 | Perceived sweetness against sucrose, independent of how hard the mix freezes |
| Stabiliser | 0.3 to 0.5% | Binds free water so ice crystals cannot grow during storage |
Quick reference. Total solids and PAC are the two that decide whether a no machine gelato works. Solids below 36 percent will be icy no matter how you freeze it, and PAC outside the band will give you either a brick or a puddle.
Most home gelato recipes fail on this table before they reach a freezer, and they fail in the same three ways every time. They measure by volume, so the solids land wherever the cup happened to land. They contain no milk powder, so MSNF sits near 8 percent instead of 10 and the base has nothing binding its water. And they use one sugar, usually sucrose alone, so sweetness and freezing point are locked together and cannot be adjusted independently. Fix those three and a no machine gelato improves more than any change of technique will deliver.
Two of those numbers are worth a second look before you weigh anything. Total solids is the anti ice number: water that is not bound by solids is water that will freeze into crystals big enough to feel, which is the mechanism behind ice crystal size and texture. And PAC is not a taste measurement at all. It counts dissolved sugar particles, so a mix with the right PAC has roughly the right proportion of its water still unfrozen at serving temperature. Get PAC wrong and no freezing method saves you.
The Base: 1000 Grams, Every Gram Accounted For
This is the site's canonical white base, which every flavour is built on. Weigh it, do not measure it by volume.
| Ingredient | Grams | Role |
|---|---|---|
| Whole milk, 3.5% fat | 575 | Water, milk fat, MSNF |
| Heavy cream, 35% fat | 165 | Fat for body and melt |
| Sucrose | 145 | Primary sugar, structure and sweetness |
| Dextrose, anhydrous | 40 | Freezing point control, less sweet per gram |
| Skim milk powder | 46 | MSNF without added water |
| Maltodextrin DE19 | 25 | Solids at PAC 0, no sweetness |
| Stabiliser blend, locust bean gum and guar | 4 | Water binding, slower melt |
| Total | 1000 |
Run the checks and it closes: fat 7.8 percent, MSNF 10.4 percent, sugars 18.5 percent, total solids 39.5 percent, PAC 278. Every figure sits inside the band in the table above, which is the point of weighing rather than guessing.
The process up to freezing is the same with or without a machine. Blend the dry ingredients into the sucrose first so the stabiliser and the milk powder disperse instead of clumping. Heat the liquids to about 40 degrees, add the dry blend, then take the mix to 85 degrees for two minutes, or hold 65 degrees for thirty, which is the low temperature route covered in the pasteurisation deep dive. Chill it fast, ideally in an ice bath rather than by leaving it on the counter.
The recipe scales linearly, so 500 grams is every figure halved, but do not scale below about 400 grams for the tray methods. A thin layer of mix in a large tray freezes solid before you can work it, and a small volume loses temperature to the room faster than you can beat it.
Then age it. Four to twelve hours at 4 degrees lets the proteins hydrate and the fat partially crystallise, and maturazione is the cheapest quality improvement available to anyone making gelato at home. Skipping it costs more texture than any freezing method will give back.
Five Ways to Freeze It Without a Machine
Churning does three jobs at once: it freezes fast, it breaks the ice crystals as they form, and it folds in a little air. Every method below sacrifices one of those, and the scored comparison of the five works through exactly where each one breaks.
| Method | Hands on time | Texture | Best for |
|---|---|---|---|
| Scrape and beat | 4 to 6 sessions of 2 minutes | Very good | The default, if you are home all afternoon |
| Salt and ice bath | 20 to 30 minutes continuous | Best of the five | Small batches, when you want it in an hour |
| Shallow tray in the freezer | 5 minutes, then wait | Fair | Sorbetti and low fat bases |
| Frozen cube blender | 10 minutes | Good but soft | Fruit bases, served immediately |
| Semifreddo | 15 minutes, then wait | Different product | When you want to skip freezing dynamics entirely |
Scrape and beat is the method to learn first. Pour the aged base into a shallow metal tray, no deeper than three centimetres, and freeze. Every forty minutes, pull it out and break the frozen edges into the liquid centre with a fork or a whisk, hard enough to grind the crystals rather than merely stir them. Four to six sessions gets you there. The shallow tray matters more than the beating: depth is what controls freezing speed.
Salt and ice bath is the fastest and gives the finest crystals. Nest a metal bowl in a larger bowl of crushed ice mixed with coarse salt at roughly three parts ice to one part salt, which drops the bath to about -10 degrees, and stir the base continuously. It is genuinely a hand cranked batch freezer without the crank.
Shallow tray alone is the least work and the coarsest result, and it is honest to say so. It suits sorbetti, where there is no fat to protect and the sugar load is doing most of the work anyway.
Frozen cube blender means freezing the base in an ice cube tray, then blitzing the cubes in a powerful blender until they come together. It gives a good texture immediately and a poor one an hour later, so serve it as you make it.
The depth rule deserves its own line, because it is the variable most people get wrong and the one that costs the most. Heat leaves a tray through its surface, so a base spread two centimetres deep freezes roughly four times faster than the same base six centimetres deep in a tub. Fast freezing produces small crystals. Slow freezing produces crystals you can feel. A wide, shallow metal tray beats a deep plastic container every time, and metal beats plastic again because it conducts rather than insulates. Chilling the empty tray in the freezer first buys another few minutes of speed at no cost.
Semifreddo is not gelato and should not be sold as gelato. Whipped cream and meringue are folded into the base so that air, not churning, provides the texture, and it freezes soft because of the aeration rather than the balance. It is an excellent dessert and a different one.
Whichever you choose, harden the result at -18 for at least two hours, then move it to the warmest part of the freezer or leave it at room temperature for five to ten minutes before serving. A home freezer runs near -18, so a tub straight from it is at ice cream temperature, not gelato temperature, and the difference in the spoon is dramatic. This is the whole content of the home freezer against a blast chiller comparison, and the practical answer at home is patience.
The Three Failures, and What Actually Causes Each
Almost every disappointing batch is one of three, and each has a single dominant cause.
Icy. Crystals you can feel on the tongue. The cause is almost always total solids below about 36 percent, which leaves too much free water, compounded by slow freezing that lets the crystals that do form grow large. Fix the recipe before you change the method, and read creamy not icy for the full diagnosis.
Gummy. A chewy, elastic, slightly slimy body that pulls away from the spoon. This is almost never the freezing. It is too much stabiliser, or a stabiliser that was never properly hydrated, and why gelato turns gummy covers both cases. At 0.3 to 0.5 percent a stabiliser blend is invisible. At 0.8 percent it is the dominant texture.
Melts too fast. A scoop that collapses into liquid on the plate within a few minutes has either too little fat, too little MSNF, or too much of its water unfrozen because PAC overshot. Why gelato melts fast separates the three, and the cure is different for each.
There is a fourth, less common and worth naming because people misdiagnose it as iciness: a dry, gritty sensation that is sandy texture, caused by lactose crystallising out of a base that carries too much milk powder. It feels like ice and is fixed by lowering MSNF, not by freezing faster.
Diagnose in that order and you will rarely be wrong: check the solids, then the stabiliser, then the PAC. It is the reverse of what instinct suggests, because instinct blames the freezer, and the freezer is almost never the problem. A batch that came out icy from a shallow tray would also have come out icy from a machine, just slightly less so.
Once the white base works, the flavours are variations on it rather than new recipes. Fior di latte is the base itself. Pistachio and hazelnut add nut paste and give back an equal weight of cream. Chocolate and stracciatella move both fat and sugar at once. Lemon sorbet leaves the dairy out entirely and rebalances from scratch. All 171 recipes on the site are balanced against the same eight numbers, and the free balancing app recalculates them the moment you change a gram.
