Are fizzy drinks ultra-processed? Almost always, yes — cola, lemonade, tonic and energy drinks are the textbook NOVA group 4 example, and the diet and zero versions typically sit in exactly the same group. Most are formulations of water, a sweetening system, an acid and flavourings, often with a colour added, and with little or no whole food anywhere in them. Plain sparkling water with nothing added is the exception: that is group 1.
Most posts in this series involve a genuine judgement call. This one is mostly useful for the second half of the answer, because a lot of people have quietly swapped to sugar-free on the assumption it is a step away from processed food. Under NOVA, it is not.
Are fizzy drinks ultra-processed? The short version
The NOVA classificationsorts food by the extent and purpose of industrial processing rather than by nutrients. Group 1 is unprocessed or minimally processed food. Group 2 is culinary ingredients such as oil, salt and sugar. Group 3 is processed food — group 1 foods with group 2 ingredients added, usually to preserve them. Group 4 is ultra-processed: industrial formulations built mostly from substances derived from foods, held together by cosmetic additives.
The fizzy drinks aisle tends to sort along that line like this:
- Cola, lemonade, tonic, energy drinks:typically NOVA 4. Water that has been sweetened, acidified and flavoured, often coloured too — check the ingredients.
- Diet, zero and sugar-free versions: also typically NOVA 4, because non-sugar sweeteners are themselves a group 4 marker.
- Plain sparkling water: NOVA 1. Carbonated water, nothing else.
- Flavoured sparkling water:depends on the label. A splash of real juice reads differently from “flavourings” and a sweetener.
Why soft drinks are NOVA's archetype
When Monteiro and colleagues describe group 4, soft drinks are one of the first examples they reach for, and the reason is structural. Ask what whole food is in a cola and the honest answer is roughly none. The ingredient list is typically water, a sweetening system, an acid and a flavouring, often with a colour — ingredients that mostly exist to make a formulation palatable rather than to preserve or cook a food.
This is the distinction we drew in UPF versus processed food: processing is not the problem NOVA is pointing at. Tinned tomatoes are processed. Cheese is processed. A soft drink is something else — a product assembled from refined components, where colour and flavour are added back because the base has neither. Preservatives, antioxidants and stabilisers on their own do not make a food group 4; cosmetic additives like flavourings, colours and sweeteners typically do, and a fizzy drink usually carries several.
It is worth being honest that NOVA coders disagree more often than the framework's confidence suggests. When Braesco and colleagues asked panels of 159 and 177 evaluators to classify two food lists in 2022, agreement was only fair — Fleiss kappas of 0.32 and 0.34. Our own reading is that agreement tends to be weakest on borderline foods, and a drink whose whole ingredient list is water, a sweetening system, an acid, a flavouring and perhaps a colour gives coders less to disagree about.
Diet, zero and sugar-free: same group
Here is the part that surprises people. Swapping a full-sugar cola for a zero-sugar one strips the free sugars out of the drink, and that is a real change to the nutrition label. It does not move the product out of NOVA 4. Non-nutritive sweeteners — aspartame, acesulfame K, sucralose, steviol glycosides — are listed in NOVA's own definition as markers of ultra-processing, on the same footing as flavourings and colours. That is about the job they are doing in the formulation — imitating sugar in a drink with no food in it — rather than about whether you can buy a box of them, because you plainly can.
So a diet drink is a formulation that has replaced one industrial sweetening system with another. If your goal is fewer free sugars, the swap does something measurable. If your goal is less ultra-processed food, it does not. Those are two different goals, and conflating them is one of the more common mistakes we see.
What about sweetener safety?
This is contested ground and we are not going to pretend otherwise. Approved sweeteners have been assessed as safe at typical intakes by food safety regulators, which is why they are permitted in the first place. Separately, health authorities have published guidance advising against relying on non-sugar sweeteners as a way to manage weight. That is a statement about whether they work as a weight strategy, not a finding that they cause harm.
We are not claiming sweeteners are dangerous, and no one should read a NOVA 4 verdict as saying so. NOVA is describing how a product was built. Whether a given sweetener is good, bad or neutral for you is a separate question that the classification is not equipped to answer.
The rest of the chiller
Plain sparkling water is group 1 and stays there — carbonation is a physical process, not an ingredient. Juice is usually read as group 1 when it is nothing but juice, though implementations differ on from-concentrate and reconstituted versions — we go into that in is orange juice ultra-processed. Either way it carries free sugars. Squash and cordial typically land in group 4 once flavourings, colours or sweeteners appear on the label, so check the pack, since a few are juice and sugar only. Kombucha varies enormously; some are group 1 or 3 ferments, others are sweetened and flavoured formulations.
The pattern is the one described in how to spot ultra-processed food: look at what the ingredients are there to do. A flavouring, a colour, a non-sugar sweetener or an industrial syrup is typically what moves a drink into group 4, because each is standing in for something the drink does not otherwise have.
What to check on the label
- The sweetening system: sugar, glucose-fructose syrup, aspartame, sucralose, acesulfame K or steviol glycosides. Plain sugar is a group 2 culinary ingredient on its own, but the non-sugar sweeteners and industrial syrups are typically group 4 markers.
- Flavourings:listed simply as “flavourings” or “natural flavourings” — both are cosmetic additives under NOVA.
- Colours: caramel colour and the rest. Present to make the liquid look like something.
- A short list with nothing added: carbonated water, full stop. That is group 1.
Or skip the reading entirely. Scanning the barcode with Cull's UPF scanner shows the NOVA verdict for a drink alongside its additives and an overall score, which is quicker than squinting at a can in the meal deal queue.
Does NOVA 4 mean a fizzy drink is bad for you?
NOVA describes processing, not nutrition, so the two questions need separating. On processing, fizzy drinks are typically group 4, and that is about as settled as NOVA verdicts get. On nutrition, there is a distinct and much older point: a full-sugar can is a large dose of free sugars in liquid form, which UK dietary advice has recommended cutting back on for years, entirely independently of anything NOVA says.
The wider evidence deserves care. Lane and colleagues' 2024 umbrella review in the BMJ, covering around 9.9 million participants, reported associations between higher ultra-processed food intake and a range of adverse outcomes, with certainty of evidence varying considerably across them. Associations from cohort studies are not proof of cause. The UK's advisory committee, SACN, set out its position in 2023 and published a rapid evidence update on 2 April 2025; it found the associations concerning, but said limitations in the evidence mean it remains unclear whether processing itself is the driver rather than these foods simply being high in energy, saturated fat, salt and free sugars.
The practical reading is undramatic. Sugared or sweetened, a fizzy drink is typically ultra-processed either way; plain sparkling water is not; and sugar-free is a sugar decision rather than a processing one. Knowing which is which is information, not an instruction.
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, BMJ umbrella review
- SACN (2023) — Position statement on processed foods and health
- SACN (2025) — Processed foods and health: rapid evidence update