The standard, and a gap between it and the law
A garden trampoline is regulated as a toy. In Great Britain that means the Toys (Safety) Regulations 2011, with the General Product Safety Regulations 2005 behind them; in Northern Ireland the European toy safety directive applies instead, and general product safety there moved onto the European regulation on 13 December 2024.
The relevant standard is BS EN 71-14, trampolines for domestic use. Its current edition is 2025, published on 31 October 2025, which on the same day superseded the 2018 edition that almost every piece of British advice still cites. Compliance is voluntary, but it carries legal weight, because meeting a designated standard gives a presumption of conformity with the regulations.
There is a gap worth knowing about. As at February 2026 the list of designated standards for Great Britain still names the 2018 edition, four months after the standards body withdrew it. So the edition a manufacturer is measured against in law and the edition a reader will find on sale as current are not the same document. That will presumably be tidied up, and until it is, no single version can be called the legal standard without qualification.
What the standard does establish is useful, and some of it is not widely known:
- It is written for one user at a time. That is in the scope, not buried in a warning — the whole document describes a product intended for a single person.
- An enclosure is required on medium and large trampolines, which means anything with a bed of 1.5 metres or more, and those models must not have a handrail. The enclosure material has to allow an adult to see the child through it.
- Padding is required, covering at least the frame and the springs, and it has to pass an impact test.
- The centre of the mat has to be marked. Aiming for the middle is a design feature, not folk wisdom.
- Size categories are set by bed diameter, with the smallest limited to a user of twenty-five kilograms, and there is a required warning where a trampoline is not intended for children under three.
Gymnastic trampolines, inflatables, fitness rebounders and anything in a public playground are all outside it, as are trampolines sold with tents or basketball hoops attached.
Trampoline parks, which are a different proposition
Parks arrived in Britain around 2014 and multiplied quickly — roughly sixty by early 2016, and about double that a decade later. They are not covered by the domestic standard and are not licensed.
The specification written for them, PAS 5000, was withdrawn on 1 December 2022 and replaced by BS EN ISO 23659, which is simultaneously the British, European and international standard. It covers design, construction, inspection and maintenance, together with minimum operational requirements, and explicitly excludes outdoor parks. It is voluntary, and in 2024 only a small number of operators were publicly known to comply with it.
Legally, a park is premises rather than a product: the Health and Safety at Work etc. Act 1974 and the law of occupiers' liability. One point is routinely got wrong — the enforcing authority is local authority environmental health, not the Health and Safety Executive, under regulations that allocate recreational premises to councils. In practice that means no routine inspection regime: councils have no duty to inspect until an accident is reported.
Where parks differ from gardens is severity rather than frequency. A review pooling eleven studies and well over a million injuries found that, compared with home trampolines, injuries at commercial centres carried higher odds of musculoskeletal injury, substantially higher odds of a lower-limb injury, and higher odds of needing surgery — while carrying lower odds of concussion, lacerations and upper-limb injury. A British series of seventy-one people brought by ambulance from one park found that nearly two-thirds of the fractures were in the lower limb, a quarter of patients were admitted, and a fifth had surgery. In a garden the pattern inverts and the upper limb dominates.
Frequency, though, still sits at home. Of the trampoline injuries in one English emergency department study, 69 per cent happened at a private address, and regional ambulance figures show call-outs to parks as a small fraction of trampoline call-outs overall.
Where injuries actually happen, and to whom
This is where the old advice misdirects, and the correction is the most useful thing on the subject.
In a prospective study of 131 children attending one English emergency department over a summer, 68 per cent were injured without falling off the trampoline at all, and the average number of people on it at the moment of injury was 2.6. In the largest mechanism dataset available — more than eighteen thousand emergency attendances across fourteen years in Victoria, Australia — falls from the trampoline accounted for about 45 per cent, falls onto the mat or frame about 19 per cent, collisions with another person about 6 per cent, and contact with the springs and frame 4.8 per cent. Over the period, falls off declined as a share while landings on the mat rose, and injuries involving more than one user rose by about a fifth each year.
So the springs are a small part of the problem, and the claim that limbs are severed in the frame appears in no British or European source at all.
On numbers and ages, the clearest British series comes from a paediatric trauma centre in Leeds, which reviewed every trampoline-related imaging request over six years:
- Attendances rose 118 per cent, from 218 in 2012 to 476 in 2017, with the largest single jump between 2014 and 2015.
- Of 788 children with positive findings there were 622 fractures and 183 soft-tissue injuries. The commonest fracture sites were elbow, wrist, ankle and knee, in that order.
- Children under seven were more likely to be injured, and the highest proportion of positive scans was at age six.
- The winter share of injuries rose from 13 to 23 per cent across the period, which the authors put down to indoor venues.
Elsewhere, an enhanced data pilot in Oxfordshire found trampolining was the commonest source of sports injury among girls attending accident and emergency, at 12 per cent, and fourth among boys; peak age was nine for boys and eight for girls. During the 2020 lockdown one Essex hospital saw trampoline cases rise from under 6 per cent of its paediatric injuries to nearly 22 per cent, and from 8 per cent of its paediatric surgical cases to 28 per cent — with parents, when asked, most often describing more than one child on the trampoline, usually with an older one.
A figure in wide circulation — thirteen thousand trampoline injuries treated in England each year — is an extrapolation published in 2015 from the Oxfordshire pilot's 2012 to 2014 data, scaled on the assumption that one county represents a particular fraction of the country. It is still being reprinted as a current national statistic, including by an NHS trust leaflet dated 2026. It is an estimate, and an old one.
At the severe end the picture is reassuring and should be stated as such. A regional paediatric neurosurgery centre recorded 44 trampoline-related cranial or spinal injuries in eleven years: nineteen cranial, nine at the craniovertebral junction, six elsewhere in the spine. Seventy per cent were managed without surgery. There was one death, an eleven-year-old boy with a bleed on the brain after striking the trampoline, which had more than one user on it. Head injuries clustered under five and spinal injuries over eleven.
The one rule with real evidence behind it
Every British source says one person at a time, the European standard assumes it, and unlike most of the advice on this subject it has a demonstrated mechanism.
A Scottish group modelled two people bouncing out of phase and found that kinetic energy transfers from the heavier to the lighter. For an adult of eighty kilograms and a child of twenty-five, the transfer is equivalent to the child falling 2.8 metres onto a solid surface, and the child's bones and ligaments also take that load faster than the adult's do. The observational data line up: an average of 2.6 people on the trampoline when a child was hurt, and multiple use as the mechanism parents themselves reported most often.
The corollary is one that a hospital leaflet draws and the national advice tends not to: an adult should not bounce with a child. The weight difference that makes it fun is the weight difference that does the damage, and the adult can also simply land on the child.
Two widely quoted figures on this — that 60 per cent of injuries occur with more than one person on the trampoline, and that the lighter person is five times more likely to be hurt — are worth attributing carefully. They circulate as research findings, but they trace back through a newspaper report to a study that does not appear to contain them. The underlying rule is well supported; those two numbers should be credited to the safety organisation that publishes them rather than to any paper.
Netting, padding and supervision: what they do, and do not, do
All three are sensible and two of them are required. None of them has been shown to reduce injuries, and saying so is more useful than repeating the recommendation.
- Padding is required by the standard and addresses a mechanism that causes under 5 per cent of injuries. An Australian analysis found the padding requirement made no statistically significant difference to spring and frame injuries, and a consumer test of twelve models found only one met it.
- Enclosures target the largest single mechanism, falls from the trampoline, and are required on anything with a bed of 1.5 metres or more. Falls off have declined as a proportion of injuries over the period netting became standard. But the studies that looked directly for a reduction in the injuries nets and pads should prevent did not find one, and netted trampolines have been associated with more multiple-user injuries — plausibly because a net keeps several children in rather than keeping one child safe.
- Supervision is advised by every source and has never been shown to work. The safety organisation that recommends it says in the same breath that it does not guarantee safety and that more than half of accidents happen under supervision — a claim that traces to an American study of 1998 — while its own briefing document states, two pages away, that almost half of injuries happened without an adult present. No study with a usable denominator exists in either direction.
What supervision can plausibly do is enforce the things that are evidenced: one at a time, no somersaults, and stepping on and off rather than jumping.
Somersaults and flips are the mechanism most often associated with neck injury, and British advice is that domestic trampolines are not suitable for them. In the one British study of a trampoline park, both documented neck injuries happened during somersault attempts. Nobody has quantified the risk, so the honest statement is an association rather than a rate — and the point applies even to a child who stays in the centre of the mat, since the injury comes from landing on the head rather than from reaching the edge.
One small contradiction, since a reader will meet both versions: on ladders, the main British safety organisation advises removing them when the trampoline is not in use, so that a small child cannot climb on unsupervised, while the Northern Ireland government's guidance suggests a ladder is worth having so that children step down rather than jump. There is no injury evidence either way.
Age, and the advice no British body gives
The firmest British position is that trampolines are not suitable for children under six. The safety organisation that says so goes further — not at all under three, and between three and six only on a trampoline built for that size — and the Northern Ireland government's guidance sets the same threshold of six. A hospital leaflet published in 2026 is softer, saying only that a younger child should use a trampoline designed for their age.
The standard itself draws its one hard age line at three years rather than six, and the British injury data support the concern about the youngest users: under-sevens were more likely to be hurt in the Leeds series, the youngest children accounted for half the cranial injuries in the Birmingham one, and in Australia multiple-user injuries among the under-fives more than doubled. What does not exist anywhere is an injury rate per child or per hour of use, for any age — so there is no way to say how much more likely an injury is, only that the burden falls disproportionately on the youngest.
One absence is worth stating plainly, because American material circulates here and reads as though it were guidance. The American Academy of Pediatrics advises against recreational trampoline use altogether and against any use by children under six. No British paediatric body takes that position, and none appears to have issued a statement on trampolines at all. The strongest British advice is the under-six threshold above, which is not the same thing as advice against trampolines.
For safety at this age more broadly, including the shift from removing hazards to handing over judgement, see health and safety. Physical activity for children covers what children of this age need by way of movement, of which a trampoline is one perfectly reasonable source.