Is your home affected by railway noise?

Where to check — and the limits of the maps

If you're buying or renting near a railway line, the good news is that you don't have to guess. Defra's strategic noise maps cover the major railways across England, giving you a reasonable starting picture of noise exposure before you've set foot in the house. The less good news is that "reasonable starting picture" is doing some work in that sentence — the maps have real limits, and it's worth knowing what they are before you treat them as gospel.

Defra's strategic noise maps are built from modelled data covering major railways, updated periodically as part of the UK's obligations under the Environmental Noise Directive. They're a genuinely useful first check — free, official, and based on real traffic volumes rather than guesswork.

But there are gaps. Coverage focuses on the busiest, most significant rail routes, so a quieter branch line or a freight spur that doesn't meet the mapping thresholds may simply not appear — silence on the map isn't the same as silence on the ground. The maps are also built on averaged, modelled traffic patterns, so they don't capture the timing of individual freight movements: a line that's mapped as moderately noisy on average can still deliver a genuinely disruptive freight train through at 2am, because Lden and Lnight are averages across a day and a year, not a guarantee about any specific service.

It's also worth being clear that these maps measure airborne noise, not vibration. Rail vibration — the low-frequency rumble and structure-borne shake you can sometimes feel in a house near track — is a separate physical phenomenon with its own assessment methods, and it isn't captured by noise mapping at all. If a property backs onto a line with heavy freight or older jointed track, vibration is worth asking about separately.

What the WHO says about railway noise specifically

The WHO's 2018 Environmental Noise Guidelines treat railway noise as its own category, with its own recommended thresholds — around 54 dB Lden and 44 dB Lnight, both carrying WHO's "strong" evidence rating. Those figures sit a little higher than the equivalent road-traffic thresholds (53 dB Lden, 45 dB Lnight on the daily average, though rail's Lnight recommendation is actually slightly tighter), reflecting a genuine, repeatedly observed pattern in the research: at an equal measured sound level, people tend to report less annoyance from rail noise than from road noise during the day. Acoustics researchers sometimes call this the "railway noise bonus."

That bonus shouldn't be overstated, though, and it doesn't apply evenly across the day. The same intermittent character that makes rail noise less annoying on average during waking hours is exactly what can make it more disruptive at night — a sudden, discrete pass-by is more likely to cause a measurable sleep-disturbance event than the same acoustic energy spread continuously, which is part of why the Lnight guideline for rail is actually very close to the road figure rather than meaningfully looser. In practice this means a railway-adjacent property can feel more tolerable day-to-day than its raw decibel figure alone might suggest, while still being genuinely worth checking carefully for its night-time exposure specifically.

Why Lden alone can mislead — check Lnight too

Lden (day-evening-night) is a weighted average across the whole day, which makes it good for comparing overall exposure but not so good at flagging an infrequent, loud event. A handful of night freight movements can be a big deal for sleep even if they barely move the Lden number, because they're diluted across 24 hours of averaging. This is exactly what Lnight is for: it isolates the night-time period specifically, so always look at both figures together rather than relying on Lden alone. A property with an unremarkable Lden but an elevated Lnight is telling you something the headline figure won't.

This gap between the two figures tends to be wider for rail than for road, precisely because rail traffic is more unevenly distributed across the 24-hour cycle — a line that's near-silent for long stretches and then carries a freight service at 3am will show a bigger Lden/Lnight divergence than a road with a more continuous, if lower, traffic flow throughout the day. If a report on a rail-adjacent property shows the two figures sitting close together, that's itself informative: it suggests a fairly even spread of movements across the day rather than a concentration of disruptive night services.

Rail noise isn't road noise: CRN vs CRTN

Road traffic noise is assessed using CRTN (Calculation of Road Traffic Noise), which models a broadly continuous stream of vehicles. Rail uses a different method entirely — CRN (Calculation of Railway Noise) — because the physics and the experience are different. CRN accounts for track geometry (curves, gradients, cuttings and embankments all change how sound propagates), the difference between welded continuous rail and older jointed track (jointed track is generally noisier, with that characteristic clatter), and the mix of freight versus passenger services, since freight trains are typically longer, heavier and noisier per pass. The upshot: rail noise tends to be about distinct, individual pass-by events rather than the continuous background rumble you get from a busy road, which is why looking only at road-style averages can undersell what living near a railway actually feels like.

No barrier or terrain modelling is applied to rail noise in AcoustaPLOT's engine at this time — unlike the road-noise pipeline, which does model terrain and building screening, rail assessments use the CRN formula itself, which already accounts for distance and track geometry but doesn't add a separate physical barrier calculation on top. That's worth knowing if you're specifically weighing up whether an existing fence, wall or embankment between a property and the railway will help: CRN's own calculation already reflects typical ground conditions, but a bespoke barrier-design calculation of the kind AcoustaPLOT offers for road noise isn't currently part of the rail assessment.

Renting vs buying: does the bar need to be different

Whether a given result matters enough to act on reasonably depends on how long you're planning to be there. A short-term rental is a more reversible decision than a house purchase — if a railway line turns out to be more disruptive than expected, moving on at the end of a tenancy is a far smaller undertaking than selling a home you've just bought. That's not a reason to skip checking before renting, but it is a reason the bar for "good enough to proceed without checking further" can reasonably sit a little lower for a short let than for a purchase you intend to stay in for years.

For a purchase specifically, it's worth remembering that an infrequent but loud event — one memorable freight train passing at an inconvenient hour — can matter more to long-term satisfaction with a home than a lower, steadier hum most people stop consciously noticing within a few weeks of moving in. Checking the actual Lnight figure, and understanding whether a line carries freight at all, is a cheap way to find out which situation you're actually looking at before committing to years of living with the answer.

Questions worth asking beyond the noise maps

A few specific questions tend to be more useful than a general "is the railway noisy" when you're assessing a property in person or asking an estate agent. Does the line carry freight, and if so roughly how often — freight services are typically longer, heavier, and noisier per pass than passenger trains, and are more likely to run through the night. Is the track welded continuous rail or older jointed track — jointed track produces a distinct rhythmic clatter that continuous welded rail doesn't. Does the property sit in a cutting, on an embankment, or at grade relative to the line — a cutting can shield a property significantly, while an embankment can do the opposite by putting the track closer to eye level with upstairs windows.

It's also worth asking, or checking via the relevant rail operator or local plan, whether any electrification, capacity enhancement, or additional service pattern is planned for the line — a quiet branch line today can see a real change in traffic volume if it's earmarked for increased freight or passenger capacity, and that kind of change won't show up in a snapshot noise assessment based on current traffic.

Roughly what different services sound like

It helps to have a rough, honest sense of scale rather than treating every "railway nearby" the same way. A local stopping service on continuous welded rail, heard from a typical back-garden distance of 20–30 metres, is often genuinely unremarkable — noticeable when it passes, forgotten a minute later. A mainline intercity service at similar distance is louder and faster, with a shorter but more intense pass-by. Freight is the outlier: longer trains mean a longer duration of elevated noise per pass, often at a lower-frequency, more penetrating pitch that's harder for a wall or window to exclude than higher-frequency passenger-train noise, and freight timetables regularly include through-the-night movements that passenger timetables don't.

None of this replaces an actual figure for a specific address — track geometry, train speed, speed restrictions near stations, and the mix of services on a given line all vary enough that a general rule of thumb about "what a train sounds like" is exactly the kind of guesswork a proper assessment is meant to replace. But it's a useful sense check when a report comes back: if the result looks surprisingly high or low for what you'd expect from what you can see from the garden, that's often a sign a line carries more freight, or runs closer to kerbing speed restrictions, than it first appears.

Check your specific address

Strategic maps give you the national picture; they can't tell you what a particular house, on a particular stretch of track, actually experiences. If you're weighing up a property near a railway, get a proper address-level check rather than relying on a national map alone. AcoustaPLOT turns Defra's noise mapping and the relevant rail-specific data into a plain-English risk assessment for your exact address, from £34.99.

Check your address at acoustaplot.com.