Usually, yes. If your windows face a busy road, higher floors often have less fresh traffic pollution, particularly nitrogen dioxide and soot. But a sheltered ground-level window can be cleaner than one much higher above the road. And for PM2.5, the difference between floors may be small.
Picture two apartments in the same building. One is near the ground and overlooks a courtyard. The other is several storeys up, with a view of busy traffic. Which would you choose for cleaner air?
Helsinki offers a surprising clue. Beside one busy road, researchers measured 27% less nitrogen dioxide and 30% less PM10 at breathing height in a sheltered courtyard than beside the road-facing wall, 14 metres (46 feet) above the ground. Same building, different side. The lower outdoor measurement was cleaner.
How can the air be cleaner closer to the ground?
Helsinki Region Environmental Services (HSY) studied Mäkelänkatu in 2017–2019. The buildings lining this road form a street canyon: traffic pollution can linger between the façades, while the blocks partly shelter the courtyards behind them.
HSY set roadside concentrations at pedestrian height to an index of 100. At the same road-facing wall 14 metres up, the indices dropped to 82 for nitrogen dioxide (NO₂) and 67 for PM10 particles. Behind the building, at breathing height, they were just 60 and 47. Comparing that courtyard with the elevated roadside point gives:
Courtyard versus the road-facing façade at 14 m
Courtyard versus the road-facing façade at 14 m
Mike’s Balance recalculation from HSY’s rounded outdoor indices: (60 ÷ 82 − 1) × 100 and (47 ÷ 67 − 1) × 100. This is not a comparison of two occupied flats.
But this is not a contest between two actual apartments. The instruments were outdoors, and HSY did not report a comparable reading high above the courtyard. That higher courtyard point might have been cleaner still. Both moving up and moving away from traffic helped in the measurements. We cannot isolate which would matter more for a particular home.
When is a higher floor actually better?
For fresh traffic emissions, often. In Bologna, Italy, researchers compared air at heights from 2 to 65 metres. Winter nitrogen dioxide fell by 74% between the lowest and highest points. Ultrafine particle counts dropped by roughly 30%.
Yet PM2.5 mass fell by only 4% in the warmer campaign and 11% in the colder one. That distinction matters more than it sounds: fresh exhaust is concentrated near its source, while much of the fine-particle mass can arrive from elsewhere in the city or region. Moving upstairs may take you away from a traffic plume without escaping the wider haze.
The same split appeared inside New York homes: black carbon was roughly 15% lower in the highest floor group than in the lowest, a statistically significant difference. But PM2.5 barely changed. These were different homes, so the study cannot tell us what would happen if the same family moved upstairs.
In short, if both apartments overlook the same busy road, being higher is often an advantage for fresh road pollution. It is not a guarantee about every pollutant or the air indoors.
Why a low window can still win
Back to our two apartments. The upper one may escape some traffic emissions simply by being farther away. But the lower one may have something else in its favour: the building itself, standing between its windows and the road.
At another Bologna site, researchers counted up to 3.4 times as many ultrafine particles outside the road-facing entrance as behind the building. The figure concerns particle number, not PM2.5 mass, but the contrast is a reminder that two sides of one building may offer very different outdoor air.
And pollution does not always decline neatly from bottom to top. In two Singapore apartment blocks beside an expressway, PM2.5 mass peaked at the middle of three sampled heights, while the highest levels were cleanest. Copenhagen and Milan studies found further departures from a simple floor-by-floor pattern, although both had measurement limitations. The details are in the study table below.
Ozone adds another twist: it increased with height in a Chicago high-rise study, while measurements in a Shenyang street canyon found roughly 40% more at lower and middle heights than at the roof. There is no universally cleanest floor for every pollutant.
What matters once you close the front door?
A cleaner patch of air outside a window does not necessarily mean cleaner air in the bedroom. Cooking, tobacco smoke, candles and poorly maintained ventilation can change indoor particle concentrations. Nor does a window take in the average air of the neighbourhood: it admits what happens to be moving past that particular opening.
In a computer model of a ten-storey block, average traffic-tracer concentrations inside rooms were approximately halved between the lowest and highest floors under one wind arrangement. Yet, under another, one ground-level room had 1.42 times the average street-level concentration, because the air outside that window was dirtier than the street average. These were modelled tracer concentrations, not observations in occupied homes.
In winter, air rising through leaky shafts and corridors can sometimes carry cooking odours or smoke between flats. In sealed-window towers, the outdoor-air intake and the building’s filters may matter more than the bedroom’s floor. And keeping windows permanently shut is not a solution: homes still need adequate ventilation.
Which apartment should you choose?
There is no honest way to pick between our two imaginary apartments without knowing more about the building. But there is a useful order of questions when you visit a real home:
- Both windows face a busy road? Going higher often helps with fresh traffic-related pollution, especially NO₂ and soot.
- One faces a sheltered courtyard? Do not dismiss the lower floor. Compare where the windows face, not just how high they are.
- The surroundings are relatively free of traffic? The benefit of moving up may be smaller, particularly for regional PM2.5. Look for other pollution sources such as chimneys, loading areas and restaurant extractors.
- The windows do not open? Find out where the ventilation system draws air from and how it is filtered.
For a close call, outdoor measurements near the relevant façades and careful indoor checks would tell you more than an estate agent’s floor number. A basic PM2.5 monitor, on its own, cannot tell you what is happening to nitrogen dioxide or ozone.
Higher is often better against fresh traffic pollution. But which side of the building you live on, and how air enters your home, can change the answer.
Every property listing tells you the floor. Very few tell you where the air outside the windows has been.
What the studies actually measured
These studies examined different buildings, pollutants, measurement periods and instruments. The table is a guide to that evidence, not a universal ranking. Heights are clearer than floor labels for international comparisons: in Britain the first floor is above the ground floor; in much of North America the first floor is at street level.
| City / study | Pollutant | How it was assessed | Finding |
|---|---|---|---|
| Helsinki · HSY KAILA (2017–19) | Outdoor NO₂; PM10 | Passive NO₂ sampling and HSY field monitoring | Road façade improved at 14 m; sheltered courtyard at ground level was cleaner still |
| Bologna · façade comparison (2016) | Ultrafine particle number | Direct front/rear particle monitoring | Front entrance: up to 3.4× the particle count at the rear |
| Bologna · height study (2018) | PM2.5 mass; UFP number; NO₂ | Instruments at five outdoor heights, two short campaigns | Lower at 65 m: PM2.5 4–11%; UFP ~30%; winter NO₂ 74% |
| New York · Jung et al. (2011) | Indoor/outdoor BC; PM2.5; PAHs | Two-week monitoring in homes; floor-group comparison | Median indoor BC: 1.59 µg/m³ on floors 0–2 versus 1.35 on floors 6–32 (significant); excluding tobacco-smoke homes, floor-group p = 0.006. PM2.5: 13.1 versus 12.8 µg/m³; outdoor middle-floor peak not significant |
| Singapore · Kalaiarasan et al. (2009) | Outdoor PM2.5 mass | Particulate sample collection for mass and chemical analysis | Middle sampling levels highest in both blocks; highest levels cleanest |
| Copenhagen · Frederickson et al. (2024) | Outdoor PM2.5; NO₂; O₃ | Five low-cost sensor nodes plus PALM simulation | Sensors: no detected vertical gradient; model varied by wind and stratification |
| Milan · Yu et al. (2025) | Indoor/outdoor PM2.5 estimates | ~30 low-cost sensors, two fortnight-long campaigns | Lower levels slightly higher outdoors on average; top floors lower I/O ratios; October exception |
| Santiago et al. (2022) · idealised city | Non-reactive traffic tracer | CFD room model; wind-tunnel evaluation of flow | Modelled mean indoor/street ratio: 0.74 on lowest versus 0.36 on highest floor under perpendicular wind; one ground-level room reached 1.42 under an oblique wind |
| Chicago · Azimi et al. (2018) | PM and outdoor O₃ | Direct vertical monitoring | Outdoor particles decreased and ozone generally increased in a one-week campaign |
| Shenyang · Miao et al. (2022) | Outdoor O₃ | One year of monitoring at three heights; street-canyon modelling | Ozone at lower and middle levels about 40% higher than at roof level |
Sources, calculations and limitations
The central Helsinki comparison comes from HSY’s KAILA monitoring summary. With roadside breathing-height concentration set to 100, the published outdoor indices were NO₂ 82 and PM10 67 at the road-facing façade 14 m up, versus 60 and 47 at breathing height in the courtyard. We calculated (60/82 − 1) × 100 = −26.8% and (47/67 − 1) × 100 = −29.9%. The input values were rounded by HSY, so the calculated percentages should not be mistaken for additional measurement precision.
The front-versus-rear contrast comes from Zauli Sajani and colleagues’ Bologna façade study (2016); the pollutant-specific height comparison from their separate vertical monitoring study (2018). Other outdoor counterexamples are Copenhagen (2024), Singapore (2009) and Milan (2025), Chicago (2018) and Shenyang (2022). The indoor New York results are from Jung and colleagues (2011); the open-window results are from Santiago and colleagues’ computational model (2022).
Important limits: The studies differ in buildings, pollutants, observation periods, sample sizes and instruments. Outdoor air at a façade is not a measurement of an occupant’s exposure or health risk. Helsinki’s courtyard comparison changes both façade and height, while the New York black-carbon association, although statistically significant, comes from different homes rather than a controlled floor-change experiment. New York’s suggested outdoor middle-floor peak was not statistically significant. Copenhagen combined low-cost sensor observations with simulations; the 40–50% gradients came from model scenarios. Santiago modelled a passive tracer in an idealised building array and did not directly validate room concentrations. Singapore sampled PM2.5 mass, not particle number. The Milan campaign was short and relied on low-cost particle sensors. Chicago was monitored for one week, whereas the Shenyang ozone series covered a year. We have not pooled these heterogeneous results into a universal safest-floor ranking or presented unpublished preliminary analyses as verified findings.
