Flavoured Water Had the Most Microplastics in This Study

If you assume plain bottled water is the cleanest option, one 2026 analysis complicates the picture. Researchers testing drinking water on Madeira Island found that flavoured and carbonated bottled waters carried the highest average microplastic load of any category they measured — higher than tap water, and more than three times higher than plain mineral water.[Molecules, 2026]

The same study found that children’s estimated exposure to those particles was roughly 3.6 times that of adults, driven purely by the math of body weight and water intake.[Molecules, 2026]

None of this proves harm. But it does show that the container and its contents matter, and that “bottled” is not a single, clean category.

What the study actually measured

The researchers analysed 22 drinking-water samples from Madeira Island: 10 bottled (four natural mineral, six flavoured or carbonated) and 12 tap waters collected across four municipalities. Particles were screened by stereomicroscopy and then confirmed chemically using micro-Fourier transform infrared spectroscopy (µ-FTIR), the technique the study used to distinguish genuine polymers from look-alike organic debris.[Molecules, 2026]

That distinction matters. Of 428 total particles recorded, only 65 were confirmed microplastics. Another 223 were non-plastics — cellulose alone made up 40% of that fraction — and 140 were classed as indeterminate, meaning identification confidence fell below 70% or the particle was unsuitable for analysis.[Molecules, 2026]

This is a useful reminder when reading microplastic headlines: a raw particle count is not a microplastic count. Rigorous chemical confirmation typically shrinks the number substantially.

Flavoured water topped the list

Concentrations across all samples ranged from 0.5 to 6 microplastics per litre. The averages fell in a clear order: flavoured water (2.00 ± 1.83 MPs/L) came in highest, followed by tap water (1.30 ± 0.80 MPs/L), with mineral water lowest (0.59 ± 0.37 MPs/L). The single most contaminated sample was a flavoured bottled water at 6 MPs/L; one tap water sample contained no detectable microplastics at all.[Molecules, 2026]

The authors attribute the elevated flavoured-water levels to extra processing and packaging steps — flavouring, mixing, syrup dosing, and multilayered linings — where mechanical abrasion or leaching from plastic components can introduce particles. They note this is consistent with earlier reports of higher microplastic loads in carbonated and flavoured beverages compared with plain water.[Molecules, 2026]

Bottled and tap water carried different plastics

The two water sources differed not just in quantity but in kind. Bottled waters were dominated by polyethylene terephthalate (PET), polypropylene (PP), and polyethylene (PE) — the polymers used in bottle bodies, caps, and sealing rings, respectively. In mineral waters, PET ranged from about 50% to 57% of identified polymers.[Molecules, 2026]

Tap waters looked different, characterised mainly by PE and PP, often alongside polyester, PET, or polyamide (PA). The authors link this pattern to distribution networks, treatment plants, and domestic plumbing, with polyester and PA fibres commonly traced to laundering and textile degradation.[Molecules, 2026]

That difference showed up in particle shape too. Fragments were the main form in bottled water — 77.8% in mineral water on average — while tap water skewed toward fibres, at a mean of 52.4%. Fragments are thought to come from the breakdown of larger plastic items and abrasion of fittings; fibres are more often attributed to synthetic textiles.[Molecules, 2026]

Across every water type, the particles were mostly small and mostly transparent. Particles under 400 µm accounted for 100% of those in mineral water, and transparent particles made up between 56% and 75% depending on the source — consistent, the authors note, with the degradation of clear PET bottles and colourless synthetic fibres.[Molecules, 2026]

Why children’s exposure ran higher

To translate concentrations into human exposure, the researchers calculated an Estimated Daily Intake (EDI), using an ingestion rate of 2.2 L/day for a 70 kg adult and 1.8 L/day for a 16 kg child.[Molecules, 2026]

The result: EDI spanned 0.01–0.19 MPs/kg/day for adults and 0.05–0.68 MPs/kg/day for children — roughly 3.6 times higher for children. The reason is straightforward. EDI is normalised to body weight, and children weigh far less while still drinking substantial volumes of water. The worst-case flavoured sample (6 MPs/L) produced the highest child EDI in the dataset at 0.68 MPs/kg/day.[Molecules, 2026]

The honest limitations

This is where good research earns trust by stating what it cannot claim.

The study’s measured concentrations were, by the authors’ own account, lower than those reported in many other countries. Earlier work has estimated adult EDIs ranging from roughly 0.31 to 13 MPs/kg/day — well above what was found here. So this is not a “bottled water is alarmingly contaminated” story; if anything, the levels were comparatively modest.[Molecules, 2026]

It is also worth being clear about what “exposure” does and does not mean here. The study estimated how many particles a person might ingest. It did not measure any health outcome. The authors note that while microplastics have been detected in human blood, placenta, and other tissues, their toxicokinetics remain incompletely understood, and the health impact of chronic low-level intake is still under investigation.[Molecules, 2026]

The practical takeaway

Read against its own caveats, the study supports a few measured conclusions rather than a scare.

First, packaging and processing appear to add particles: the more a water is handled — flavoured, carbonated, multilayer-packaged — the more microplastics it tended to carry in this dataset. Plain mineral water was consistently lowest.[Molecules, 2026]

Second, tap water is not automatically cleaner than bottled. It carried its own signature of plumbing- and textile-derived plastics, and one sample was clean while another exceeded most bottled samples.[Molecules, 2026]

Third, the study’s authors’ central call is not a consumer instruction but a systemic one: standardised monitoring protocols, better production standards, and rigorous risk assessment of chronic low-level exposure — reinforced, they argue, by the EU’s move toward a harmonised methodology for measuring microplastics in drinking water.[Molecules, 2026]

For an individual, the reasonable reading is modest: this is a real, measurable, low-level exposure that varies with the container and the processing behind it — worth tracking as the science matures, not worth panicking over on the strength of 22 samples.