WHO Environmental Noise Guidelines explained

The headline figures

Noise reports are full of numbers, and numbers on their own don't tell you much unless you know what they're being measured against. The World Health Organization's 2018 Environmental Noise Guidelines for the European Region give us that reference point — a set of figures that turn "this road is 58 dB" from a meaningless statistic into something you can actually assess.

For road traffic, the WHO guidelines give two key values: around 53 dB Lden and around 45 dB Lnight. Lden is a day-evening-night weighted average, giving you a sense of overall daily exposure; Lnight looks specifically at night-time levels, which matters because sleep disturbance doesn't average out the way daytime annoyance does. Above these guideline levels, adverse health effects — and specifically sleep disturbance in the case of Lnight — become more likely at a population level. That's the number to hold in your head when you're looking at a noise report: not a hard line between "fine" and "dangerous," but a threshold above which the odds shift.

Road traffic isn't the only source the guidelines cover, and the thresholds aren't identical across sources, which is worth knowing if you're comparing a road-noise report against a railway one. Railway noise gets a slightly higher recommended threshold — around 54 dB Lden and 44 dB Lnight — reflecting evidence that, at an equal measured level, rail noise tends to be somewhat less annoying than road noise during the day, even though it can still disturb sleep at a lower level than daytime annoyance alone would suggest. Aircraft noise sits at the strictest end, around 45 dB Lden and 40 dB Lnight, because aircraft noise has consistently shown stronger links to annoyance and cardiovascular effects at a given measured level than either road or rail. Every one of these three recommendations carries WHO's "strong" rating — its highest confidence category — meaning the evidence was judged robust enough to recommend policy action, not just flagged for further study.

What Lden and Lnight actually measure

Lden and Lnight sound like jargon, but they describe something fairly intuitive once unpacked. Lden is a day-evening-night average: it takes the noise over a full 24 hours but applies a weighting penalty to evening noise and a larger one to night-time noise before averaging, specifically because the same traffic noise is more disruptive at 11pm than at 11am. The metric is built to reflect that reality rather than treating every hour of exposure as equally harmless.

Lnight is simpler: it's purely the average noise level during night-time hours, with no daytime data diluting it. This is the figure most directly linked to sleep disturbance, which is where a large share of the WHO's health evidence concentrates. Neither metric is the same as the single loudest moment you might remember from a property viewing — one lorry downshifting outside, say, or a particularly noisy train. They're long-run averages, which is exactly why a desktop assessment built from traffic-flow data can estimate them meaningfully, while a ten-minute visit at the wrong (or right) time of day often can't.

Why WHO set these thresholds at all

It would be easy to assume these figures are about comfort — the noise-equivalent of "some people prefer it quieter." They're not. The WHO guidelines were developed as health-protective thresholds, based on evidence linking environmental noise exposure to effects including sleep disturbance, cardiovascular strain, and impacts on wellbeing. The guidelines exist because environmental noise is now recognised as a genuine public health issue, not merely a quality-of-life annoyance — which is precisely why they're framed as levels to stay below, rather than aspirational targets.

The strongest evidence base behind the 2018 guidelines concerns three outcomes specifically. Sleep disturbance comes in two forms: the kind you notice, where noise actually wakes you, and the kind you usually don't, where noise fragments the structure of sleep without a conscious awakening — both are captured in the evidence behind the Lnight threshold. Annoyance is treated as a health outcome in its own right, not just an inconvenience, because sustained annoyance has its own measurable link to stress and wellbeing. And cardiovascular effects describe a chronic, cumulative risk — exposure above these thresholds over years is associated with a measurably higher risk of effects like hypertension, not an acute one-off reaction to a single loud night.

A simple way to self-assess: comfortably below vs well above

You don't need to be an acoustician to use these figures sensibly. The most useful way to think about them isn't precise arithmetic — it's a comfortably below vs well above framing. A property that sits comfortably below both the Lden and Lnight guideline values is unlikely to be a noise problem in any meaningful sense. A property that sits well above them, particularly on Lnight, is one where noise-related sleep disturbance is a real and reasonable concern, worth factoring into your decision. It's the properties in between — hovering close to the line — where a proper assessment earns its keep, because that's exactly where "fine" and "not fine" become hard to tell apart by ear alone.

As a worked example: a semi-detached house set back from a B-road, with a measured or modelled exposure of around 48 dB Lden and 38 dB Lnight, sits comfortably below both WHO thresholds — a result most people would be entirely happy living with. A terraced house directly fronting a busy A-road junction, by contrast, might come back at 65 dB Lden and 52 dB Lnight — well above both, and squarely in the range where the WHO's own evidence says sleep disturbance becomes a reasonable concern, not a hypothetical one. Most real UK properties, inevitably, land somewhere between these two examples, which is exactly the point of getting an actual number rather than guessing from a viewing.

What these numbers are — and aren't

One caveat worth holding onto: these are population-level guideline values, drawn from evidence about how groups of people respond to noise exposure on average. They are not a precise, individual verdict on your specific health or sleep at a specific address. Individual sensitivity to noise varies considerably — some people sleep through a busy road, others are woken by a distant train — so treat the WHO thresholds as a well-evidenced steer, not a certificate. They tell you where the odds shift, not what will definitely happen to you personally.

It's also worth being precise about the shape of that shift. Noise harm doesn't switch on at 53.0 dB and off at 52.9 — the underlying evidence describes a dose-response relationship, where risk rises gradually as exposure increases. The WHO's thresholds mark the point where the evidence for population-level harm becomes strong enough to recommend acting on, not a hard boundary that separates a dangerous level from a safe one. A property at 56 dB Lden isn't in a different category to one at 50 dB Lden so much as a few steps further along the same gradually rising curve.

This is also why two honest, well-informed people can look at the same figure and reach different personal conclusions about whether to buy a specific property — one treating a result just above the Lnight guideline as a dealbreaker, another treating it as manageable with better glazing. Both can be reasonable responses to the same number; the guidelines describe population-level risk, not a verdict on any one household's tolerance.

Where this fits against UK planning and building standards

It's worth being clear that the WHO guidelines are a health recommendation, not a UK legal requirement — they don't automatically become a planning condition or a building regulation. UK planning policy on noise is actually built around a separate framework: the Noise Policy Statement for England sets out three effect levels — a No Observed Effect Level below which there's no detectable impact, a Lowest Observed Adverse Effect Level above which effects first become noticeable, and a Significant Observed Adverse Effect Level above which development is normally considered unacceptable unless strong countervailing benefits apply. For internal noise inside a dwelling specifically, British Standard BS 8233 sets the relevant benchmark figures UK planning conditions typically reference.

These UK-specific frameworks are informed by much of the same underlying evidence the WHO guidelines summarise, but they're not a line-by-line transposition of the WHO numbers into planning policy, and a property can satisfy a UK planning noise condition without necessarily sitting under every WHO threshold, or vice versa. That distinction matters practically: a WHO-based noise risk band tells you about likely health and wellbeing impact, calibrated against the best available international evidence. It isn't a claim that a property has or hasn't met a specific UK legal or planning noise standard — that's a separate question, assessed against a separate document, and worth checking on its own terms if it matters for your situation (our companion piece on new-build noise and planning covers that framework in full).

Check where your address actually sits

None of this is meant to turn a house viewing into a health risk assessment. The practical use of these thresholds is comparative, not diagnostic: they let you understand roughly where a property's noise exposure sits relative to a well-evidenced reference point, rather than relying on how a ten-minute viewing happened to sound on the day it took place.

The figures only become useful once you know how a real address compares to them, rather than a hypothetical one. AcoustaPLOT takes Defra's strategic noise mapping and DfT traffic data and works out exactly where a specific UK address sits against these WHO thresholds, in plain English, from £34.99.

Check your address at acoustaplot.com.