Bottled water is sold on an idea: that it is the clean option, the careful choice, the one you make when you are looking after yourself. It is worth asking what is actually in the bottle, because the research of the last few years has turned up things that were not in the marketing.
This article sticks to what has been published and what it does and does not prove. There is a lot of frightening material written on this subject, much of it going well beyond the evidence. We would rather give you the real picture, including the parts that are reassuring.
The 2018 study that started the conversation
In 2018, the journalism organisation Orb Media commissioned researchers at the State University of New York in Fredonia to test bottled water properly. They analysed 259 individual bottles from 11 leading brands, bought in 19 locations across 9 countries, including several in the Middle East and Asia.
Plastic particles turned up in 93% of the bottles tested. The average was 325 particles per litre, and the spread was enormous — some bottles showed effectively none, while one contained more than 10,000. The plastics identified included polypropylene, nylon and PET, the material most bottles are made from. The pattern pointed at the bottling process and the packaging itself rather than the water source.
The finding was significant enough that the World Health Organization announced a review. That review, published in 2019, is the part that usually gets left out, so it deserves stating plainly: the WHO concluded that microplastics in drinking water appear to pose a low risk to health at the levels then being found. Particles larger than 150 micrometres are unlikely to be absorbed by the body, and uptake of smaller ones was expected to be limited.
The WHO was equally clear that the available information was limited, that the evidence on the smallest particles was thin, and that more research was needed. It did not say microplastics were harmless. It said nobody yet had grounds to say they were harmful, which is a different statement.
What better instruments found in 2024
The gap the WHO pointed at was the very small particles — the ones older equipment simply could not see. In January 2024, researchers at Columbia University published work in the Proceedings of the National Academy of Sciences that closed some of it.
Using a technique called stimulated Raman scattering microscopy, they could count and chemically identify particles down to around 100 nanometres, far below the limit of earlier methods. In bottled water, they found an average of roughly 240,000 detectable plastic particles per litre — between 10 and 100 times more than previous estimates. Around 90% of them were nanoplastics, the fraction earlier studies had been blind to.
Two things are true about that number at the same time, and both matter.
The first is that it is not evidence that bottled water became more contaminated. It is evidence that we could finally see what was already there. The second is that nanoplastics behave differently from larger particles: they are small enough to cross biological barriers that microplastics cannot. That is precisely why the WHO flagged them as the open question in 2019.
What none of this shows is a specific health effect. Researchers have found the particles. What they do inside the body over a lifetime of exposure is genuinely still being studied, and anyone telling you otherwise — in either direction — has gone past the evidence. We are not going to claim bottled water causes any disease, because no one has shown that it does.
Tap water is not exempt
Honesty requires this bit. The same organisation that funded the bottled water work had already tested tap water in 2017, across five continents. Plastic fibres turned up in 83% of 159 samples, averaging about 4 particles per litre.
So this is not a story of dirty bottles and pure taps. Microplastics are in the environment and therefore in the water supply generally. What the two studies together suggest is a difference of scale — the bottled figures came in dramatically higher, and the packaging itself appears to be a large part of the reason.
BPA, and why “BPA-free” is not the end of the sentence
Bisphenol A is an industrial chemical used to make hard, clear polycarbonate plastic and certain epoxy resins that line containers. It can migrate out of that plastic into what it holds. It is a well-studied endocrine disruptor, and regulators have grown steadily more cautious about it: in 2023 the European Food Safety Authority cut its tolerable daily intake dramatically, and the EU banned BPA in food contact materials from January 2025.
Two clarifications are worth having.
The ordinary single-use bottle is usually PET, which is not made with BPA. The material people should actually be asking about is polycarbonate — the hard, rigid, slightly blue-tinted plastic used for many of the large returnable water jugs sat in office and home dispensers. Those are the containers where the question is live, particularly older ones that have been through many cycles of washing and refilling.
The second clarification is the important one. “BPA-free” does not mean “free of bisphenols”. Manufacturers largely replaced BPA with close chemical relatives — BPS and BPF — which laboratory work suggests can act on the body in similar ways. Chemists have a name for this pattern: regrettable substitution. Replacing a studied chemical with an unstudied one that shares its structure buys a label, not necessarily a safer product.
Heat, which is where this gets local
Most of the research above was done in temperate countries. The UAE changes the arithmetic, because everything here spends time hot.
PET is made using antimony compounds as a catalyst, and small amounts of antimony can migrate from the bottle into the water. How much depends on two things: how long it has been stored, and how warm it has been. The relationship with temperature is steep. Studies have found measurable increases at 50°C within 24 hours, and the higher the temperature goes, the faster it happens. Work looking specifically at bottled water in a Gulf climate found the same pattern.
Now think about the actual journey a bottle takes here. A warehouse that is not climate controlled. The back of a delivery van in the afternoon. A stack of six-packs in a garage or a store room off the kitchen. A bottle left in a parked car for a few hours, where interior temperatures climb far past anything the bottle was designed around.
None of that is exotic. It is the ordinary supply chain in a country where it is over 40°C for months. The same product tested in a European laboratory has had an easier life than the one in your store room.
The honest bottom line
Nobody should panic about a bottle of water. If you are handed one, drink it. The measured risks are, as far as anyone can currently tell, small, and there is no published finding that would justify alarm.
But when you set it out end to end — particles that better instruments keep finding more of, bisphenol questions that got relabelled rather than resolved, and a heat problem that this climate makes worse — the case for bottled water as the careful, premium, health-conscious option gets fairly thin. You are paying a premium for a container that is doing some of the work you were trying to avoid, and then throwing the container away.
Filtration is not magic either, and it is worth being specific about what it does. Activated carbon is what handles chlorine, taste and odour. If particle removal is the thing you care about, that is the job of a membrane: a 2019 study found reverse osmosis removed 99.9% of micro and nanoplastics from tap water, and RO remains the only household technology that reaches meaningfully below one micron — though the very smallest nanoplastics are reduced rather than eliminated. That is the accurate version, and it is worth more to you than a bigger claim.
If you want to see how the different systems compare on what they actually remove, our comparison page lays them out side by side. We have also written about what is in Dubai’s tap water before it reaches your kitchen, which is the other half of this question.