Europe's summer heat killed 50,391 city residents, with two thirds of deaths tied to climate change

Europe's summer of 2026 will be remembered less for what it looked like than for what it cost. Across 854 cities in 30 countries, researchers now estimate that heat killed 50,391 people between June and August, a toll larger than any previously recorded summer in the region and nearly twice as high as each of the two next most dangerous years on record. The figures come from two peer-reviewed papers in Nature Health, and they carry a second, harder finding: 32,619 of those deaths, or 65 percent, are attributable to human-caused climate change.

The work was led by researchers at Imperial College London and the London School of Hygiene & Tropical Medicine. Rather than counting death certificates on which heat appears as a cause, the authors modelled mortality using established relationships between temperature and death rates for different age groups, then compared the observed summer against a simulated world in which fossil fuel emissions had never warmed the atmosphere. The distance between those two worlds is the burden the study assigns to climate change, and that distance has been widening for a quarter of a century.

Europe has counted heat deaths before. The 2003 heat wave became a reference point for a generation of public health officials, and the summer of 2022, with 29,480 deaths according to the same modelling team, was until now the worst this century. What separates 2026 is not only the size of the toll but the share of it traceable to warming. At the start of the century, about 40 percent of heat deaths in the studied cities were attributable to climate change. This summer the figure stood at 65 percent, roughly two in every three.

The geography was wide. Western Europe recorded its hottest summer on record, about 2.9C above the 1991-2020 average, and some days ran more than 12C above normal. Death rates per million residents were highest in Malta, Italy and Hungary. Absolute tolls were largest in Italy, France, Spain, Germany and the UK, where the researchers estimated 4,354 heat deaths. The cities in the sample account for about 40 percent of the population of the countries studied, which are the EU-27 plus the UK, Switzerland and Norway.

Key Facts

The Guardian reported on October 23 that the toll was called staggering by the scientists who produced it, and that climate breakdown was responsible for two in every three deaths. The newspaper quoted co-author Malcolm Mistry saying that the absolute numbers were staggering but that the proportion attributable to anthropogenic climate change was even more alarming. The Guardian also reported on October 23 that the UK alone recorded an estimated 4,354 heat deaths. The modelling covered repeated heat waves in June, July and August, and because it excludes a deadly May heat wave and covers only the 854 cities in the sample, the authors describe the figures as underestimates.

The Associated Press reported on October 23 that 50,391 people died from warmer temperatures and that 32,619 died because of the extra heat caused by burning coal, oil and gas. The agency reported that the toll is 71 percent higher than the previous deadliest summer, 2022, when the same researchers estimated 29,480 deaths. It also reported that about one-third of the deaths occurred between June 17 and July 1, when temperatures were 5.3 degrees Celsius above normal, according to co-author Garyfallos Konstantinoudis of Imperial College London. Lead author Clair Barnes said some European cities saw temperatures 12 degrees Celsius warmer than normal.

CNN reported on October 22 that the summer is likely to have been the worst on record for heat-related mortality across much of Europe, quoting Barnes, an extreme weather and climate change researcher at Imperial College London. The network reported that scientists compared temperatures in 854 cities across 30 European countries with a modelled world without climate change and then applied peer-reviewed heat-mortality relationships for June through August. It noted that the toll is the highest in 26 years and substantially above both the 2022 and 2003 heat waves, and that southern and western Europe were worst affected while cooler northern regions, including the UK, still recorded thousands of deaths.

The age profile is stark. People aged 85 and older accounted for roughly half of all deaths, an expected pattern given how heat stresses ageing cardiovascular systems. But 833 people aged 20 to 44 also died, a figure that undercuts any assumption that this is a problem confined to the very old. In Naples, the hardest-hit city by rate, 815 people per million residents died from the heat, and an emergency doctor there reported a 15 percent increase in emergency room visits during the hottest weeks.

The method matters for how the numbers should be read. The team compared computer simulations of a world without climate change against observed deaths, then applied peer-reviewed heat-mortality relationships city by city. That approach yields estimates rather than body counts, and it is the same approach that produced the 2022 and 2003 comparisons, which is what makes the year-on-year figures comparable. The headline number is an estimate; the trend it describes is not.

Analysis

The single most important number in these papers is not 50,391. It is 65 percent. A death toll measures weather; an attributable fraction measures responsibility. When roughly 40 percent of European heat deaths at the turn of the century could be pinned on anthropogenic warming and 65 percent can be pinned on it now, the change is not a description of a hotter summer but of a shifting baseline in which the same thermometer reading produces more death than it used to. What this really means is that Europe's heat mortality is no longer primarily a natural phenomenon with a climate component attached; the climate component has become the majority of the phenomenon.

The comparison with 2022 is the other number that deserves scrutiny. A 71 percent increase over the previous record in four years is not a gradual deterioration. Part of that jump reflects the severity of 2026 itself: western Europe's hottest summer on record, about 2.9C above the 1991-2020 average, and days more than 12C above normal. Part of it reflects the shape of the summer, with repeated heat waves rather than a single event, and a devastating stretch from June 17 to July 1 that alone produced about a third of the deaths. But it also reflects population ageing in cities designed for a cooler climate, and the fact that housing stock, tree cover and emergency capacity determine whether a given temperature becomes a mortality event.

The geography supports that reading. Malta, Italy and Hungary recorded the highest death rates per million, while Italy, France, Spain, Germany and the UK carried the largest absolute tolls. The UK figure of 4,354 deaths is a useful corrective to the assumption that this is a Mediterranean problem. Britain is not a hot country, and its homes, hospitals and public spaces are not built for sustained extreme heat. Northern cities with little air conditioning and little experience of heat emergency planning produce death at lower temperatures than southern cities that have adapted over decades.

The bigger picture here is that heat is now the deadliest form of climate impact in the world's wealthiest temperate region, and its victims are concentrated in a demographic that every European country has more of each year. A continent with an ageing population, dense urban housing and rising summer temperatures has constructed a structural vulnerability that no single heat wave created and no single heat wave will end. The 50,391 figure is a measurement of that structure, not an accident that happened to it.

Why It Matters

It matters first because the number is an underestimate. The modelling excludes May, which was itself deadly, and covers only 854 cities representing about 40 percent of the population of the countries studied. It does not cover rural populations, smaller towns, or countries outside the EU-27, the UK, Switzerland and Norway. Every methodological caveat in these papers points in the same direction: the recorded toll is a floor, not a ceiling.

It matters second because of timing. CNN reported on October 22 that the findings were published weeks before international leaders gather in Turkey for COP31, and days after a UN report warned that the chance of limiting warming to 1.5C had vanished. A death toll of this size, with two-thirds of it attributed to fossil fuel pollution, lands in the middle of a negotiation whose central question is how fast to stop burning coal, oil and gas. Attribution studies rarely change negotiating positions, but they do change the framing: heat deaths stop being a public health footnote and become a measurable consequence of emissions policy.

It matters third because the deaths are concentrated in ways that make them tractable. People over 85, the 833 people aged 20 to 44, residents of dense southern cities such as Naples with 815 deaths per million, and an emergency department reporting 15 percent more visits during the hottest weeks: these are identifiable populations in identifiable places. Heat warning systems, cooling centres, home retrofits, tree canopy and hospital surge planning all act on the same variables the study models. That 2026 was the deadliest summer on record does not mean the deaths were inevitable; it means the defences were not built to the standard the climate has now reached.

Next Up

Attention now turns to COP31 in Turkey, where the study's authors and the wider climate research community will press the case that heat mortality should be treated as a headline indicator of climate harm rather than a secondary public health concern. The publication of two peer-reviewed papers in Nature Health weeks before the summit gives delegations a citable, city-level dataset covering 30 countries, a far more concrete basis for adaptation finance arguments than the global death estimates that circulated until now.

In the meantime, the researchers have signalled that the 2026 numbers are incomplete. A deadly May heat wave sits outside the June to August window, and the 854-city sample leaves most of the continent's territory unmeasured. Extending the analysis to further cities and to the full warm season is the obvious next step, and it will almost certainly raise the toll. The question for European governments is not whether the next estimate will be worse, but whether the defences put in place before the next summer will be sized for the climate that produced this one.