Is ice cream ultra-processed? From our own scanning of UK freezer aisles, most commercial tubs carry emulsifier and stabiliser systems plus flavourings, and typically read as NOVA 4. Homemade custard ice cream, made from milk, cream, sugar, egg yolk and vanilla, is a culinary preparation and typically sits in NOVA 3. Ice cream is also named directly in NOVA's own published examples of group 4 foods — though the ingredient-list reading below is an interpretation, not a settled fact.
What makes ice cream worth its own post is the reasoning. The usual assumption is that the sugar and the cream are what make it ultra-processed. Under NOVA they do not. Understanding what actually moves the group tells you more about the framework than almost any other food on the shelf.
Is ice cream ultra-processed? The short version
The NOVA classificationsorts foods by the extent and purpose of industrial processing, not by nutrients. Group 1 is unprocessed or minimally processed food. Group 2 is culinary ingredients — sugar, salt, oils, butter. Group 3 is processed food: group 1 foods with group 2 ingredients added. Group 4 is ultra-processed food, meaning industrial formulations built from derived substances and cosmetic additives.
Ice cream lands on both sides of that line:
- Homemade custard ice cream:milk and cream are group 1, sugar is group 2, egg yolk is group 1, vanilla is a seasoning. Churning and freezing a custard is a cooking technique. That is typically a NOVA 3 dish — the same logic that keeps a homemade cake batter or a pot of stewed fruit out of group 4.
- Many commercial tubs:the same base plus an emulsifier and stabiliser system — mono- and diglycerides of fatty acids, guar gum, locust bean gum and relatives — usually with flavourings and sometimes colours. That combination typically reads as NOVA 4.
Why sugar and cream are not the deciding factor
This trips up a lot of people. Sugar is a group 2 culinary ingredient. Adding it to a group 1 food makes the result a processed food, group 3. It does not make anything ultra-processed. The same is true of butter, salt and oil. If sugar content decided NOVA groups, homemade jam and meringue would be ultra-processed. Under the framework they typically are not.
So the sweetness is not the marker. Neither is the fat. What NOVA watches for is the presence of substances and techniques a domestic kitchen does not use: cosmetic additives such as flavourings, colours, flavour enhancers and non-sugar sweeteners, and industrial ingredients such as glucose-fructose syrup, maltodextrin, modified starches and protein isolates. That is where an industrial tub separates itself from a churn at home.
The emulsifier and stabiliser question — honestly
Here is the part that deserves care, because it is a judgement call rather than a rule read off a chart.
Under NOVA's own descriptions, stabilisers, gums, thickeners and antioxidants are group 3 additives. They do not, on their own, make a food ultra-processed. A single gum in a product is not a group 4 trigger, in the same way that a preservative in a tin does not make its contents UPF.
What tips most ice cream into group 4, on our reading, is different. Industrial ice cream does not use one stabiliser incidentally. It uses an emulsifier and stabiliser system, several ingredients working together, to do a job a kitchen recipe is not built for: hold overrun and a smooth texture through a factory freezer, a lorry, a supermarket cabinet and the partial thaw and re-freeze of a home freezer. That is a cosmetic function — building sensory qualities — and it usually arrives alongside flavourings, which are a group 4 marker in their own right.
That is the reading Cull applies, and it is consistent with how NOVA treats ice cream in its published examples. But it is an interpretation, not a settled fact. When two panels of 159 and 177 evaluators sorted two food lists against NOVA in a 2022 study, agreement between them was only fair — Fleiss kappa values of 0.32 and 0.34. That is a reminder that the boundary is not crisp in practice.
Ice cream or frozen dessert? A UK labelling quirk
There is a naming pattern on UK shelves worth knowing, though it answers a different question. From our own scanning of UK freezer aisles, some tubs are labelled dairy ice cream, which signals milk fat, while others are sold as frozen desserts, typically built on vegetable fat.
That is useful information about what you are buying. It is not a NOVA verdict. A frozen dessert carrying flavourings typically lands in group 4 — but a dairy ice cream can carry a full additive system too, so the name on the front does not tell you. Only the ingredient list tells you about the processing.
What to check on the tub
Twenty seconds on the back of the pack settles most cases:
- A short dairy list: milk, cream, sugar, egg yolk, vanilla. That reads like a kitchen recipe and typically sits in NOVA 3. In our scanning these exist but take some looking for, usually at the premium end.
- Flavourings:the word “flavouring” — as opposed to named vanilla, cocoa or fruit — is a cosmetic additive and a typical group 4 marker.
- Colours: another cosmetic additive, and a clear signal of an industrial formulation.
- An emulsifier and stabiliser system: mono- and diglycerides, guar gum, locust bean gum, carrageenan and similar, used together. Individually these are group 3 additives; as a system building texture and overrun, they are what tips most tubs into group 4 on our reading.
- Industrial ingredients: glucose-fructose syrup, maltodextrin, modified starch, whey or milk protein isolate. These are group 4 markers on their own.
Or skip the reading. Scanning a barcode with Cull's UPF scannershows the NOVA verdict for that product along with its additives and an overall score — helpful in a freezer aisle where the interesting text is on the underside of the lid.
Nutrition is a separate check
This matters, and it cuts both ways. NOVA describes processing, not nutrition. A homemade custard ice cream can land in group 3 and still be high in free sugars and saturated fat — a NOVA 3 verdict says nothing about either, and those sit on the nutrition panel, worth reading separately. Equally, ultra-processed and unhealthy are not synonyms: the evidence links higher UPF intake to adverse outcomes with varying certainty, which is not the same as any one tub being harmful.
The honest state of the evidence supports that caution. A 2024 BMJ umbrella review covering roughly 9.9 million participants reported associations between higher UPF intake and a range of adverse health outcomes, with varying certainty across those outcomes. The UK's advisory committee, SACN, has looked at this twice — a 2023 position statement and a rapid evidence update published on 2 April 2025 — and its position remains cautious: the associations are concerning, but limitations in the research mean it is unclear whether processing itself is the cause, rather than these foods simply being high in energy, saturated fat, salt and free sugars.
Ice cream is a useful illustration of that ambiguity, in the way yoghurt is at the other end of the chiller. Both are dairy; both split into a plain version that typically stays out of group 4 and a flavoured version that usually does not. In both cases the group tells you about the processing, and the nutrition panel is the separate check.
So the practical reading is modest. If you want less ultra-processed food in your freezer, look for the short dairy ingredient list and accept that it will cost more and melt faster. If you would rather make it, a custard base is four ingredients and a churn. And if the tub with the additive system is the one you actually enjoy, that is a choice you are entitled to make with the information in front of you — which is all a NOVA group ever was.
Sources
- Monteiro et al. (2016) — NOVA. The star shines bright, World Nutrition
- Open Food Facts — NOVA groups for food processing
- Braesco et al. (2022) — Ultra-processed foods: how functional is the NOVA system? European Journal of Clinical Nutrition
- Lane et al. (2024) — Ultra-processed food exposure and adverse health outcomes: umbrella review, BMJ
- SACN (2023) — Position statement on processed foods and health, and the SACN rapid evidence update published 2 April 2025