Working at height: the regulations behind the leading cause of construction fatalities

Falls from height are consistently one of the leading causes of fatal and serious injury in UK construction, which is part of why the Work at Height Regulations 2005 are written the way they are: broadly, and without a minimum height threshold. A common misconception is that the regulations only apply above two metres, or some other specific figure — they don't. They apply to any work where a person could fall a distance liable to cause them personal injury, which in practice can mean a stepladder, a loading bay edge, or a shallow mezzanine just as much as it means a roof.

Why 'no minimum height' matters in practice

This framing is deliberate rather than an oversight. A fall from a low platform onto a hard surface, or awkwardly onto an obstruction, can cause a serious injury even though the fall itself looks minor from the outside — and site teams who mentally file "working at height" under "anything involving scaffolding or roofs" tend to under-control the lower-height tasks that don't feel dangerous in the moment. A worker standing on a stepladder to run a cable, or briefly on a low stack of pallets to reach something, is doing work at height in the regulatory sense, even if nobody on site would naturally describe it that way.

The hierarchy: avoid, prevent, mitigate

The regulations set out a clear order of preference for managing height risk, and it's worth actually following the order rather than jumping straight to whichever option is most convenient. First, avoid work at height altogether where it's reasonably practicable to do so — assembling something at ground level before lifting it into position, for instance, rather than assembling it in place at height. Where height work genuinely can't be avoided, the next step is preventing a fall from happening at all, through collective protection like guardrails and edge protection that don't depend on an individual worker doing anything correctly to stay safe. Only where a fall can't be prevented does the hierarchy move to mitigating its consequences — nets, airbags, or a properly used harness system — which is a real, useful layer of protection, but one that assumes a fall might still happen rather than stopping it happening in the first place.

Where this hierarchy gets skipped in practice

The most common failure isn't ignoring the hierarchy outright — it's jumping straight to the mitigate step because it's the most familiar and often the quickest to arrange. Handing someone a harness and lanyard feels like taking working at height seriously, and it is one genuine part of managing the risk, but if the same task could have been avoided entirely by changing the sequence of work, or prevented with guardrails that were available but not used, relying on the harness alone means accepting a fall might still happen when a better option existed. A RAMS for height work that jumps straight to PPE without addressing whether the work could have been avoided or prevented first hasn't genuinely worked through the hierarchy — it's picked the easiest-sounding control and stopped there.

Equipment and inspection

Where height work equipment is used — scaffolding, mobile access towers, harnesses, ladders — it needs to be suitable for the specific task, properly maintained, and inspected on a schedule appropriate to how it's used and how quickly its condition can change. This overlaps directly with routine site inspection discipline: a scaffold inspected thoroughly and recorded properly is a genuinely different risk profile to one that's been walked past and assumed fine. Ladders in particular are worth a specific mention, because they're often treated as low-risk by default simply because they're familiar and used constantly — a ladder is a legitimate option for short-duration, low-risk tasks, but it sits low in the hierarchy of controls, and using one for a task that genuinely warranted a tower or fixed platform is a common way the hierarchy gets quietly skipped.

What happens when a fall protection system does its job

A fall successfully arrested by a harness, with no injury resulting, generally isn't itself RIDDOR-reportable under the current rules — no injury occurred, and dangerous occurrence categories are narrowly defined. This is worth being aware of, though, because it's specifically the kind of event the proposed 2026 RIDDOR changes are looking at extending reporting to cover, on the logic that an unplanned fall-arrest activation is meaningful evidence a fall very nearly happened, independent of whether the system caught it. Regardless of whether that specific proposal is eventually adopted, an unplanned fall-arrest activation is worth treating as a serious near miss internally right now — it's telling you something genuinely went wrong with a control that was relied on to prevent a fall, and that's worth investigating on its own merits, not just when a regulation eventually requires it.

A worked example

A roofer's lanyard catches them after a slip near a roof edge — no injury, no fall to the ground, the system worked exactly as intended. Under a narrow reading, nothing reportable happened and the day continues. Under a genuinely useful safety approach, that activation gets logged and reviewed the same way any other near miss would be: was the edge protection adequate, was the anchor point correctly positioned, was the roof surface wetter or more slippery than the RAMS anticipated. Treating a caught fall as "nothing happened" because the outcome was fine misses that the underlying hazard — whatever caused the slip — is still present on that roof for the next person working up there.

Weather and changing conditions

Height work is more sensitive to changing conditions than most other construction tasks, because wind, rain and ice all directly affect the risk in ways that a RAMS written days in advance can't fully anticipate. It's worth having a genuine, followed process for reassessing height work when conditions change during the day — not just a line in the RAMS that says "do not work at height in high winds," but an actual point of authority on site who checks conditions and can stop height work if they've deteriorated since the task was planned. This is the same real-time judgement that matters for permit-to-work systems on other high-risk tasks: conditions checked in advance aren't the same as conditions right now.

Competence for height work specifically

The regulations also require that anyone working at height, and anyone supervising or organising that work, is competent to do so — which for height work specifically tends to mean genuine, task-relevant training in the equipment being used, not a general awareness of safety principles. A scaffolder is trained specifically in scaffold erection and inspection; that training doesn't automatically transfer to competence in assessing a mobile elevating work platform, even though both fall under the broad umbrella of "height work." It's worth checking that competence claims are specific to the actual equipment and task involved, in the same way subcontractor competence generally needs to be assessed against the specific job, not a general reputation.

Common mistakes

• –Assuming the regulations only apply above a specific height threshold, when there is no minimum height in law
• –Jumping straight to PPE-based fall protection without genuinely considering whether the work could be avoided or prevented first
• –Using a ladder for a task that, on reflection, warranted a more robust access solution higher up the hierarchy
• –Treating a caught fall or fall-arrest activation as a non-event because no injury resulted
• –Not having a real process for reassessing height work when weather conditions change during the day
• –Assuming competence for one type of height-work equipment transfers automatically to another