European cities have reduced many air pollutants over several decades. That progress is real, but calling London, Paris, or Stockholm a “clean-air success” without defining the pollutant and benchmark is misleading.
The European Environment Agency (EEA) reports that concentrations have improved across EU Member States, yet 94% of the EU urban population remained exposed to PM2.5 above the World Health Organization's 5 µg/m³ annual guideline in its latest assessment. The figure and method are published in the EEA Air Quality Status Report 2025.
The accurate story is therefore progress with a large unfinished health gap. London, Paris, and Stockholm illustrate different policy tools and different kinds of evidence.
The Benchmarks Matter
Three benchmarks are often mixed together:
- WHO guideline: health-based guidance; annual PM2.5 is 5 µg/m³.
- Current legal limit: the enforceable standard under national and EU law for the relevant year.
- 2030 EU limit: the revised Ambient Air Quality Directive sets an annual PM2.5 limit of 10 µg/m³ from 2030, according to the European Commission's Zero Pollution dashboard.
A city can comply with today's legal limit and still exceed the WHO guideline. “Legal” and “health-risk free” are not synonyms. The European Commission says the revised directive brings 2030 standards closer to WHO recommendations, not that it makes them identical; the new rules entered into force in December 2024.
London: A Vehicle Policy With an Evaluated Counterfactual
London's Ultra Low Emission Zone (ULEZ) charges vehicles that do not meet specified emissions standards. It is primarily a road-transport policy, so nitrogen dioxide and exhaust emissions are central outcomes. It should not be presented as a complete solution to all PM2.5, which also comes from heating, industry, regional transport, secondary formation, non-exhaust road wear, and other sources.
The most useful feature of London's evidence is that the official evaluation does more than compare “before” and “after.” The Greater London Authority models a no-ULEZ counterfactual and reports uncertainty around attribution. Its London-wide ULEZ One Year Report estimates that in 2024 roadside NO2 was 27% lower across London than in a scenario without any ULEZ phases. The estimates were 54% lower in central London, 29% lower in inner London, and 24% lower in outer London.
Those are modeled policy impacts, not direct statements that measured pollution fell by exactly those percentages. The same report separates general trends from ULEZ attribution: 99% of the 92 monitoring locations used in its 2019–2024 comparison improved, but not every improvement can be assigned to the zone.
The report also estimates that PM2.5 exhaust emissions from cars and vans in outer London were 31% lower than expected because of the London-wide expansion. “Exhaust” is an important qualifier. Brake, tire, and road wear remain, and ambient PM2.5 includes non-traffic sources.
What London demonstrates
- Define the regulated vehicle standard and geographic boundary.
- Publish monitor selection and a counterfactual method.
- Separate measured trends from modeled policy effects.
- Report pollutant-specific outcomes instead of a generic AQI improvement.
Paris: Measured Progress, But Not a Completed Transition
Airparif, the official regional air-quality observatory, reported 2024 urban-background PM2.5 of approximately 8–9 µg/m³ in Paris. The existing legal limit was met across the city, but the WHO annual guideline was exceeded throughout Paris. Both statements appear in Airparif's 2024 Paris assessment.
This is a better description than “Paris now has clean air.” The concentration was below the new 2030 EU limit in much of the city, while locations close to busy roads remained higher and the WHO benchmark was not met.
Paris and the Greater Paris metropolitan area use several transport policies with different purposes:
- Since January 2025, the low-emission zone inside the A86 has included restrictions on Crit'Air 3, 4, 5, and unclassified vehicles during specified periods. The City of Paris notes that the education period, with only occasional informational checks and no penalties, runs through 31 December 2026. The current rules and exemptions are on the city's official ZFE page.
- Since November 2024, the central limited-traffic zone has restricted through-traffic while allowing destination traffic. The city explains that this limited-traffic zone is different from the emissions-class-based ZFE.
These policies have plausible mechanisms: fewer high-emitting vehicles and less through-traffic can reduce local traffic emissions. But a policy announcement is not an outcome evaluation. We should not claim that a particular street redesign or vehicle restriction produced a fixed health benefit unless an official assessment or peer-reviewed causal study measures it.
Airparif's broader regional analysis reports a downward PM2.5 trend and emphasizes weather. The exceptionally wet and mild conditions in 2024 favored lower particle concentrations. That is a reminder that one clean year cannot be attributed entirely to policy.
What Paris demonstrates
- Publish annual concentration maps, not just an index color.
- Distinguish current legal compliance from the WHO guideline.
- Treat implementation status, enforcement status, and measured outcomes as separate facts.
- Account for meteorology when interpreting year-to-year change.
Stockholm: Pricing, Monitoring, and Stricter Future Standards
Stockholm introduced congestion charging to reduce traffic and improve access. The Swedish Transport Agency's current rules charge vehicles entering or leaving the inner city and using Essingeleden at specified times. In 2026 the maximum daily charge is SEK 135 in high season and SEK 105 in low season, as listed on the agency's official Stockholm congestion-tax page.
Those prices and dates describe the policy; they do not by themselves establish an air-quality effect. Evaluating the effect requires traffic, fleet, emissions, weather, and monitoring data.
Stockholm's environmental administration and SLB-analys publish an annual station-level report. The 2024 report, Luften i Stockholm, separately evaluates current legal standards, Sweden's stricter “Clean Air” environmental objective, and WHO guidelines. That structure is valuable because a city may pass a legal test while missing a more protective health benchmark.
Stockholm also illustrates source specificity. Road dust and studded-tire wear are important for PM10 along streets, while PM2.5 includes a larger regional and combustion-related component. A congestion charge or cleaner fleet may reduce some emissions without eliminating resuspended road dust or pollution transported from outside the city.
What Stockholm demonstrates
- Pair transport pricing with a transparent monitoring network.
- Evaluate PM10, PM2.5, and NO2 separately.
- Keep local street concentrations distinct from urban-background exposure.
- Continue planning for stricter 2030 standards even when current standards are met.
What Can Be Transferred to Other Cities?
No policy can be copied without local source analysis, but several practices travel well.
1. Build a trustworthy baseline
Use regulatory monitors, publish data coverage, distinguish roadside from urban-background sites, and report a completed calendar year in physical units. Without a baseline, a city cannot tell whether a policy worked.
2. Target the source and pollutant
Low-emission zones are most direct for vehicle exhaust and roadside NO2. Clean heating rules target building combustion. Dust management targets PM10. Regional power and industrial controls can affect secondary PM2.5. The intervention should match the emissions inventory.
3. Evaluate the counterfactual
Weather, economic activity, fleet turnover, fuel prices, and regional emissions change even without a new policy. A credible evaluation estimates what would have happened otherwise and discloses assumptions.
4. Measure distribution, not only the city mean
Roadside communities may remain more exposed after the citywide average improves. Neighborhood maps and exposure indicators reveal whether benefits reach the most affected populations.
5. Report trade-offs and enforcement honestly
Compliance rates, exemptions, financial support, enforcement dates, traffic displacement, and affordability are part of implementation. Omitting them turns policy analysis into promotion.
Do Not Compare European Cities With “Average AQI”
The EEA's real-time European index is designed for current communication. It is not the same as the U.S. AQI, and neither should be averaged into a long-term city score. For cross-city trends, use annual pollutant concentrations from the EEA European City Air Quality Viewer or official city reports, preserving the year and monitoring method.
An overall European index also cannot automatically be converted to cigarette equivalents. This site's model requires a U.S. PM2.5 AQI. If the dominant pollutant is ozone or NO2, or the index uses European breakpoints, the conversion is invalid. Use the methodology to check the input boundary.
Bottom Line
London, Paris, and Stockholm provide useful evidence about low-emission zones, traffic restrictions, congestion pricing, and transparent monitoring. They do not prove that any European city has finished the job. Europe has made substantial progress, while 94% of the EU urban population in the EEA's latest assessment remained above WHO's annual PM2.5 guideline. The transferable lesson is disciplined measurement and policy evaluation—not a celebratory ranking built from unsupported average AQI values.