ROR Labs cover: 5x more on wood fibre. A cluster randomised trial in Toronto. Arm fractures ran at 1.9 per 100,000 student-months on sand and 9.4 on engineered wood fibre.
|

The Safety Surface Under the Climbing Frame Lost to Sand, Five to One, in a Randomised Trial.

Key takeaways · 14 min read

  • 213,000+ US children a year in emergency departments for playground equipment injuries, mostly aged 5–12.
  • Climbers 36%, swings 30%, slides 20%. Three quarters are falls. Fractures are 35% of injuries and over 90% of admissions.
  • The only randomised trial of surfacing: arm fractures 1.9 per 100,000 student-months on sand against 9.4 on engineered wood fibre, from only 22 fractures in total, with borderline significance and no difference in overall injuries.
  • Impact standards test head injury from a vertical drop. The injury that fills the emergency department is a broken arm, which is a different test.

The surface under a climbing frame is the one part of a playground that exists purely for safety. Nobody plays on it; it is there to catch children. It costs a great deal, and the material chosen is usually engineered wood fibre, the pale shredded chips that meet the impact standard, drain well and look tidy.

In 2009 a cluster randomised trial compared that surface with plain granitic sand across a set of Toronto schools. Arm fractures from falls onto the surface ran at 1.9 per 100,000 student-months on sand and 9.4 on engineered wood fibre, roughly a fivefold difference, in the direction nobody expected.

That result has real caveats, because the total number of fractures involved was small, and they are set out below. It is still the only randomised trial of playground surfacing anyone has run, and it points at something worth understanding: the impact standard the industry certifies against measures head injury from a vertical drop, and most playground injuries are not that. This article covers where the 213,000 annual emergency visits actually come from, what the evidence supports, and two things (one item of safety equipment and one design instinct) that make children less safe rather than more.

A play tower with a gabled roof, a ladder and a slide standing on pale surfacing, a climbing arch to one side, low sun behind the trees.
The last hour of the afternoon, before the park empties.

Where the injuries come from

More than 213,000 children under 18 are treated in US emergency departments each year for playground equipment-related injuries, and the bulk are aged 5 to 12. The mechanism is overwhelmingly simple.

Which equipment, and what happens

Share of playground equipment-related emergency department visits, US children.

Climbing equipment36%
Swings30%
Slides20%
3 in 4of these injuries are caused by a fall — not a collision, not equipment failure.
35%are fractures. Among children admitted to hospital, fractures are more than 90%.
This is why surfacing is the variable everyone argues about. Three quarters of the injuries end with a child hitting the ground, and the ground is the one thing adults choose.

Source: national study of US playground equipment-related injuries treated in emergency departments, Center for Injury Research and Policy, Nationwide Children’s Hospital.

The shape of that second number matters. Fractures are a third of all injuries but more than nine tenths of those serious enough to admit. Stated precisely, the playground problem is broken arms.

The one randomised trial, and what it found

Andrew Howard and colleagues enrolled 37 Toronto elementary schools that were due to replace their playground surfacing anyway and randomly assigned each to engineered wood fibre or granitic sand. Twenty-eight schools completed the trial with new surfaces, covering 15,074 students. Injuries were collected prospectively over more than two years, and the primary outcome was upper extremity fractures caused by falls onto the surface.

Arm fractures from falls onto the surface

Rate per 100,000 student-months, among schools that installed their assigned surface.

1.9Granitic sand. 95% CI 0.04 to 6.9.
9.4Engineered wood fibre. 95% CI 3.7 to 21.4.
About a fivefold difference, p ≈ 0.049 — which is to say, just inside conventional significance and not a comfortable margin. The authors’ conclusion was that granitic sand reduces arm fractures compared with engineered wood fibre, and that surfacing standards should be updated to reflect it.

Source: Howard, A.W. et al., “School Playground Surfacing and Arm Fractures in Children: A Cluster Randomized Trial Comparing Sand to Wood Chip Surfaces”, PLOS Medicine, 2009.

Now the caveats, which matter

That number comes with four weaknesses, and any one of them would be enough for a reasonable person to want the trial repeated.

Read the result with these attached

Four reasons to hold this loosely

THE NUMBERS ARE SMALLTwenty arm fractures on wood fibre and two on sand. A handful of events either way would move the result across the significance line. This is a signal, not a settled fact.
NOT EVERY SCHOOL COMPLIEDOf 28 schools with new surfaces, 19 installed what they were assigned; nine crossed over to the other material. The headline figures come from the compliant schools, which is a weaker comparison than the randomisation was designed to give.
TOTAL INJURIES WERE SIMILARThe effect was specific to arm fractures from falls onto the surface. Overall playground injury rates did not separate. Sand is not a general-purpose safety upgrade.
THE SURFACES WERE WELL MAINTAINEDFracture rates came out lower than the researchers expected, which they attributed partly to unusually good upkeep. A depleted, compacted surface of either material behaves differently, and that is what most real playgrounds have by August.

The honest summary: one randomised trial, borderline significance, pointing the opposite way from industry practice. That is worth knowing and not worth ripping out a playground over.

A mechanism also helps explain why the result is not as strange as it sounds. Impact-attenuating standards are written around head injury from a vertical drop, the critical fall height at which a headform experiences forces associated with life-threatening injury. That is the right test for the worst outcome, but it does not capture what breaks an arm. A wrist arrives at the ground at an angle, with a body behind it, and what matters is how far the surface displaces under a small hard contact area. Loose sand moves. Compacted wood fibre resists.

The depth is the part that fails

A slide seen straight on from the bottom, its side rails dark against the sky, the ground beneath the run-out worn into a hollow.
The ground gives up first where everybody lands.

Whatever material is under the equipment only works at depth, and loose-fill surfacing migrates: it gets kicked out from under swings, scuffed away at slide exits, and compacted along the paths children actually run. Those are exactly the landing zones. A playground inspected in April and used all summer is a different playground in September, and nothing about the certification reflects that.

The practical takeaway that survives all the caveats above is that the maintained depth of the surface under the landing zones is a more useful thing to look at than the material. If you can see the hard base through the fill under a swing or at the bottom of a slide, the surface is not doing the job it was specified to do, whichever material it is.

The helmet that has to come off

This one is short, not widely known, and has killed children.

In 1999 the US Consumer Product Safety Commission issued a warning after a 3-year-old boy in Pennsylvania died while wearing his bicycle helmet on playground equipment. He became wedged sliding through the structure and the helmet strap strangled him. A 7-year-old girl in Canada had died the same way two years earlier, caught in an opening in a play structure. The agency also had four non-fatal cases: two children climbing trees and two on playground equipment.

The helmet is the mechanism. It turns a child’s head into a shape that passes through a gap and then does not come back out, with the chin strap already around the neck, and play structures are full of openings sized for a head without a helmet on it.

The rule, in one line

Helmet on the bike. Helmet off everywhere else.

The CPSC chair put it as plainly as it can be put: children should always wear a helmet while riding their bikes — but when a child gets off the bike, take off the helmet.

WHERE IT GOES WRONGThe bike ride ends at the park. The child runs straight to the climbing frame with the helmet still on, because taking it off is a step nobody thought about. Every documented case has this shape.
TREES COUNTTwo of the non-fatal cases were children climbing trees. Branches make the same kind of gap that play structures do.
THIS IS NOT AN ARGUMENT AGAINST HELMETSCycling helmet evidence is strong and this changes none of it. It is an argument about when. The helmet is protective on the bike and hazardous on the equipment, and the transition between those two states is the thing to build a habit around.

If you supervise a school run, a scout group or a birthday party where children arrive by bike, this is the single highest-value sentence in this article.

The other failure: playgrounds too dull to use

Everything above pushes toward making playgrounds safer, but a body of evidence pulling the other way deserves equal weight, because the risk-elimination instinct has its own cost.

A 2015 systematic review examined risky outdoor play (play at height, rough and tumble, and children moving independently without an adult in sight) across studies covering roughly 50,000 participants aged 3 to about 13. The evidence base is uneven, and the review says so: only one eligible study addressed play at height, and the strongest evidence sits under independent mobility. What it found ran consistently in one direction, though. Children with greater independent mobility were substantially more physically active; in one study, girls with high independent mobility were 4.5 times more likely to reach recommended activity levels. Rough and tumble play was associated with better motor development and social competence.

The authors name what the standards-driven playground has become “Kit, Fence, Carpet”: a certified structure, a boundary, and a rubber mat. Their objection is that it has limited appeal, and a playground children do not want to use produces no injuries and no benefit. Their framing of the trade-off is worth quoting directly: injuries are an inevitable side effect of physical activity, and physical activity is not optional for a healthy childhood.

Two errors, opposite directions

Too hardDepleted surfacing, worn landing zones, equipment above a height the surface was specified for. Produces the fractures.
Too softNothing to climb, nothing to test, an adult within arm’s reach at all times. Produces a playground nobody uses.
These are not opposites to be balanced by splitting the difference. They are different variables. Maintain the surface properly and let the play be interesting — the trade-off people assume exists between the two is much weaker than it feels.

Sources: Howard et al., PLOS Medicine, 2009; Brussoni, M. et al., “What is the Relationship between Risky Outdoor Play and Health in Children? A Systematic Review”, International Journal of Environmental Research and Public Health, 2015.

What to actually look at

A ninety-second check, in order of how much it matters

CheckWhy it is in this position
1. Helmets off before the equipmentThe only item on this list with documented deaths behind it, and the only one that costs nothing and takes one second.
2. Depth under the landing zonesUnder the swings, at the bottom of the slide, around the base of climbers. Hard base showing through means the surface is not doing its job regardless of material.
3. Height of what they are climbingFalls are three quarters of the injuries and fall height is the term that scales the energy. Equipment installed for older children sitting in a space used by four-year-olds is the common mismatch.
4. Drawstrings, scarves, necklacesSame failure mode as the helmet strap: anything around the neck plus any gap. Winter is the season for this.
5. Where you standUnder the climber, not on the bench. Supervision that prevents a fall has to be within reach of the equipment with the highest fall height, which is usually the climbing structure.
6. Whether it is interesting enoughGenuinely on the list. A playground children abandon after ten minutes has costs that do not show up in injury statistics.
Notice that nothing here is a purchase.

Sources: US Consumer Product Safety Commission helmet strangulation warning, 1999; playground injury epidemiology from the Center for Injury Research and Policy; Brussoni et al., 2015.

Nothing linked in this article, and why

We are not selling you playground safety equipment

PLAYGROUND HELMETSThey exist. Given that the documented playground helmet deaths were caused by a helmet, we are not going to recommend adding one. Nothing in the injury data suggests head protection is the missing piece on a climbing frame.
WRIST GUARDS FOR PLAYThere is real evidence for wrist guards in snowboarding, where the fall is predictable and repeated. Nobody has shown a benefit for general playground use, and the arm fracture problem is being driven by the surface and the fall height, not by unguarded wrists.
HOME PLAYGROUND SURFACING KITSDepth and maintained coverage are what matter, and a bagged product cannot supply either. If you are installing home equipment, the useful spend is on adequate volume of loose fill and topping it up, not on a branded material.
CHILD GPS TRACKERS AND WEARABLESAlready on our refusal list, and this article makes the point sharper: the review evidence supports children having more independent mobility, not more monitoring of it.
“SAFETY CERTIFIED” AS A PURCHASE SIGNALThe certification tests head injury from a vertical drop. It is a real test of a real thing. It is not a test of the injury that fills the emergency department, and a label claiming it does not tell you the surface is at depth today.

The two interventions with the best support in this article — take the helmet off, and keep the surface at depth — are both free.

A wheelbarrow standing on pale surfacing beside a heap of bark chips, the timber edging of the play area running behind.
Somebody has to top it up, and keep topping it up.

What the CPSC handbook says about playground surfacing

This section collects what the US Consumer Product Safety Commission (CPSC) handbook and checklist say about the ground under play equipment. It is general information, not a safety inspection.

Loose-fill surfacing: Table 2 of the CPSC Public Playground Safety Handbook (materials not tested to an impact standard)
MaterialInstalled depthCompressed depthCritical height listed
Rubber mulch8 in*6 in*10 ft
Sand12 in9 in4 ft
Pea gravel12 in9 in5 ft
Wood mulch (non-CCA)12 in9 in7 ft
Wood chips12 in9 in10 ft

The handbook’s note on the asterisk says rubber loose-fill does not compress the way the other materials do, but the depth should stay constant. It also says loose-fill compresses at least 25% over time through use and weathering, and that the critical height can be reduced by extreme heat or cold, wind and poor drainage. (CPSC Handbook, Pub. 325)

Other points in the handbook: concrete and asphalt must never be directly under equipment, grass and dirt are not protective surfacing, and the surfacing rating should be at least the fall height of the tallest equipment. It names ASTM standards (ASTM is a non-government standards body) for tested surfaces and cautions against untested recycled-tire rubber mulch.

The CPSC checklist, published with the nonprofit KaBOOM!, gives rules of thumb: at least 12 inches of wood chips, mulch, sand or pea gravel, or safety-tested mats; surfacing that extends at least 6 feet in all directions; swings with a clearance of twice the height of the suspending bar; and structures over 30 inches high at least 9 feet apart. The CDC adds that climbing equipment and swings are more likely to lead to concussions and recommends soft material below equipment.

What we could not confirm: the handbook’s use-zone dimensions, a general maximum equipment height, and a CDC playground injury statistics page. The handbook’s cover carries the code “092024” while its interior footers read July 2025, so we describe it as the 2025 edition with that caveat.

What to do

Steps sorted by when they matter. Each one says how strong the evidence is: Measured Official advice Untested

If it happens

  • Helmets off before the equipment. It is the only item here with documented deaths behind it. Official advice see Sources
  • Check the depth under the landing zones: swings, the bottom of the slide, the base of climbers. Hard base showing through means the surface is not working. Measured see Sources
  • Match the height to the child. Falls are three quarters of the injuries, and fall height scales the energy. Measured see Sources
  • Remove drawstrings, scarves and necklaces: anything around the neck plus any gap. Official advice see Sources
  • Stand under the climber, not on the bench. Untested see Sources

What not to do

  • Do not add a playground helmet. The documented playground helmet deaths were caused by a helmet. Official advice see Sources

Questions people ask

Should I ask our school to replace the wood chips with sand?

On one borderline trial, no. That is not enough evidence to justify the cost, and the same trial found no difference in total injuries. Maintenance is worth raising: whether the surface is being topped up to specified depth, particularly under swings and at slide exits, and who checks. That costs little, has a clearer mechanism, and applies whichever material is already installed. If a school is replacing surfacing anyway and asks which to choose, the trial is a legitimate thing to put in front of them, with its caveats attached.

My child hates having the helmet taken off at the park. Is this really a big deal?

Two documented deaths and four near-misses is not a large number, and the absolute risk to any one child on any one afternoon is very small. The item is on this list because avoiding it costs a single sentence and one second, which makes the trade-off unusually lopsided. A framing that works with children: the helmet is bike equipment, and it comes off when the bike stops, the same way a seatbelt does.

Is a rubber unitary surface better than loose fill?

It has a real advantage that has nothing to do with impact: it does not migrate, so it cannot end up below depth, and it is accessible to wheelchairs. Those are good reasons to choose it. On the injury question there is no randomised comparison against sand, so anyone telling you it is definitively safer is going beyond the evidence. What we can say is that the failure mode people actually encounter with loose fill, scattered and thin exactly where children land, is one that rubber does not have.

What about home playground equipment?

Same physics, and usually worse execution. Home sets are frequently installed straight onto grass or bare soil, both of which are poor surfaces, and the required loose-fill depth surprises people — it is measured in tens of centimetres, not a scattering. If you install equipment at home, the surfacing volume is the part of the budget most often underestimated. Grass is not a surface; it compacts within a season and is hardest exactly under the swing.

Sports and recreation injuries in general are their own subject and we will come back to them; this article is about the equipment.

How closely should I be supervising?

Close enough to the highest thing they can fall off, and no closer than that. The evidence in this article cuts both ways deliberately: a fall from the climbing structure is where the fractures come from, so that is where an adult is worth having, but the same review that documents playground injury also documents that children who move independently get substantially more physical activity. Standing under the climber is useful. Following a nine-year-old around the whole playground is not, and has costs of its own.

The short version

  • 213,000+ US children a year in emergency departments for playground equipment injuries, mostly aged 5–12.
  • Climbers 36%, swings 30%, slides 20%. Three quarters are falls. Fractures are 35% of injuries and over 90% of admissions.
  • The only randomised trial of surfacing: arm fractures 1.9 per 100,000 student-months on sand against 9.4 on engineered wood fibre, from only 22 fractures in total, with borderline significance and no difference in overall injuries.
  • Impact standards test head injury from a vertical drop. The injury that fills the emergency department is a broken arm, which is a different test.
  • Maintained depth under the landing zones beats the choice of material, and it is the thing that degrades over a summer.
  • The helmet comes off before the climbing frame. Two children have died of strangulation by helmet strap on play equipment; four more survived it.
  • A playground nobody wants to use has costs that do not appear in injury statistics. Children with more independent mobility are markedly more active.
  • Nothing on the list of things that work is a purchase.

This article summarises published injury epidemiology, one randomised trial and one systematic review. It is general information rather than a safety inspection, and it is not a substitute for the playground standards that apply where you live, which differ between countries. If a child has fallen and there is deformity, an inability to use the limb, loss of consciousness, vomiting or drowsiness, that is a matter for medical assessment rather than an article.

Further reading: Last Child in the Woods — Richard Louv (Algonquin Books, 2005). The book that put the shrinking radius of childhood on the agenda. Be warned that its central phrase, “nature-deficit disorder”, is the author’s coinage rather than a diagnosis, and the argument runs ahead of the evidence in places — read it for the framing, not for the causal claims. Find it on Amazon (paid link)

On the links above: some are affiliate links, marked (paid link). If you buy through one we may earn a commission at no additional cost to you. As an Amazon Associate I earn from qualifying purchases. We link to product searches rather than specific items so that recommendations do not break as models change, and we say plainly when we are choosing not to link something. Full policy: Affiliate Disclosure.

Sources

  • Howard, A.W., Macarthur, C., Rothman, L., Willan, A. and Macpherson, A.K., “School Playground Surfacing and Arm Fractures in Children: A Cluster Randomized Trial Comparing Sand to Wood Chip Surfaces”, PLOS Medicine, 2009. (37 schools enrolled, 28 completing with new surfaces, 15,074 students; upper extremity fractures from falls onto the surface at 1.9 per 100,000 student-months on granitic sand, 95% CI 0.04–6.9, against 9.4 on engineered wood fibre, 95% CI 3.7–21.4, p ≈ 0.049; 2 fractures on sand and 20 on wood fibre; 19 of 28 schools compliant with assignment; overall injury rates similar between groups.)
  • National study of playground equipment-related injuries treated in US emergency departments, Center for Injury Research and Policy, Nationwide Children’s Hospital. (More than 213,000 children under 18 treated annually; majority aged 5–12; climbers 36%, swings 30%, slides 20%; more than three quarters of injuries caused by falls; fractures 35% of all injuries and more than 90% of those requiring admission.)
  • US Consumer Product Safety Commission, “After Recent Death CPSC Warns Against Wearing Bike Helmets on Playgrounds”, 1999. (Death of a 3-year-old boy in Pennsylvania on 4 February 1999 and of a 7-year-old girl in Canada in 1997, both by strangulation on the helmet strap after becoming caught in play equipment; four additional non-fatal incidents, two involving tree climbing; guidance that helmets be worn while cycling and removed when the child leaves the bike.)
  • Brussoni, M., Gibbons, R., Gray, C., Ishikawa, T., Sandseter, E.B.H., Bienenstock, A., Chabot, G., Fuselli, P., Herrington, S., Janssen, I., Pickett, W., Power, M., Stanger, N., Sampson, M. and Tremblay, M.S., “What is the Relationship between Risky Outdoor Play and Health in Children? A Systematic Review”, International Journal of Environmental Research and Public Health, 2015. (21 papers from 8 countries, approximately 50,000 participants aged 3 to 12.99; strongest evidence base under independent mobility; girls with high independent mobility 4.5 times more likely to meet recommended activity levels in one study; positive associations for rough and tumble play with motor development and social competence; critique of “Kit, Fence, Carpet” playgrounds.)
  • CPSC. Public Playground Safety Handbook, Pub. 325 (PDF dated July 2025 in its footers). cpsc.gov; Public Playground Safety Checklist cpsc.gov.
  • CDC. Outdoor Play and Safety for Children in ECE, reviewed Aug. 20, 2026. cdc.gov.

Similar Posts