Document No. CSG-GDE-GRD-001 Rev. 3 Issued 2026-08-07 Next review 2027-08-07 Owner Canadian Safety Group Inc.

Field guide · Ontario construction

A guardrail is only as strong as the fasteners holding it down

Ontario tested more than 90 wood guardrails to the breaking point. Quebec has since tested more than 350 more. The wood was rarely the problem. The nails were.

Who this is for: site supervisors, carpenters, forming crews, formwork and framing contractors, and anyone who builds, checks, or removes a wood guardrail. Read time: about 10 minutes.

Read this first: where this guide applies. This guide is about temporary wood guardrails on construction projects in Ontario, under the construction regulation (O. Reg. 213/91, s. 26.3). It does not apply to permanent guardrails on finished buildings, decks, balconies, or stairs - those follow the Ontario Building Code, which is a different law with different rules. If you are building a deck rail at home, this is not your guide.

The short version

If you read nothing else, read this.

A guardrail in Ontario must hold a 150 lb (675 N) push on the top rail. That is not much. A worker who trips and falls sideways into a rail can put far more than that on it.

In the Ontario tests, guardrails built with common nails broke at 45 to 85 lb (200 to 380 N). That is a third of what the law asks for. The same guardrail design, built with screws, held 150 to 300 lb (670 to 1,330 N).

Same wood. Same size. Same shape. The only change was what held it together.

Nails do not hold a guardrail. Screws do.

The parts of a guardrail

Everyone needs to use the same words on site.

Every guardrail needs all three: a top rail, a mid rail, and a toe board. Miss one and it is not a guardrail.

The base is where guardrails actually fail, so it has its own names.

The three circled joints are where the tested guardrails came apart. All three need screws. One nailed joint drags the whole post down to the strength of that joint.

What Ontario law asks for

Construction Projects, O. Reg. 213/91, section 26.3.

You must use a guardrail when

A worker can reach the open edge of a floor, bridge deck, roof with formwork in place, scaffold platform, work platform, runway, or ramp, and could fall 2.4 m (8 ft) or more.
A worker could fall more than 3 m (10 ft) from any spot.
A worker could fall more than 1.2 m (4 ft) where wheelbarrows or similar equipment travel.
A worker could fall into machinery, water, another liquid, or onto something hazardous.
There is an opening in the floor or roof that is not covered.

Guardrails come first. Harnesses come after. You only move to travel restraint, fall restricting, fall arrest, or a net when a guardrail is not practicable.

Size and position

What	Rule
Top of the system	0.9 m to 1.1 m (35 in to 43 in)
Mid rail	Halfway up
Toe board height	89 mm (3.5 in) min
Distance back from the edge	300 mm (12 in) max
Post spacing	2.4 m (8 ft) max
Post and rail size	38 x 89 mm (3.5 in) min
Wood type	S-P-F, construction grade or better

The wood must have no visible defect that affects how much load it can carry, and no splinters or sticking-out nails.

The four strength tests

The guardrail must hold each of these loads, anywhere along its length, one at a time.

In the regulation these are written in newtons: 675 N (150 lb), 450 N (100 lb), 450 N (100 lb) and 225 N (50 lb). On site, think 150, 100, 100 and 50 lb (220 N).

What the tests found

Over 90 wood guardrails pulled to failure by the Construction Safety Association of Ontario. Each bar is the average breaking load of the end post, which is always the weakest post in a run.

Fails the 150 lb (675 N) rule Passes 150 lb (675 N) legal minimum

Steel base post, common nails to deck

45 lb (200 N)

Common nails at every joint (wood post, gusset, base)

52 lb (230 N)

Steel base post, spiral nails to deck

67 lb (300 N)

Nailed gusset, spiral nails to deck

85 lb (380 N)

Screws everywhere except the gusset, which is spiral nailed

112 lb (500 N)

Steel base post, 2 screws to deck

156 lb (695 N)

Screws at every joint, one gusset

164 lb (730 N)

Screws at every joint, one gusset (posts turned the other way *)

219 lb (975 N)

Screws at every joint, two gussets

298 lb (1,325 N)

Middle posts in the same runs held more again. The best middle post average was 366 lb (1,630 N), and one middle post reached 401 lb (1,780 N). Every bar above uses the end post number, because that is the post a worker is most likely to land against and the one you should design to.

* How the post is turned changed the result here, and Quebec testing in 2019 found the same effect. Turning a post the strong way helps. Both studies also note it is hard to control on a real site, so treat it as a bonus, not a substitute for screws and two gussets.

Four things the numbers show

Fasteners decide the strength. Of everything tested, how the parts were fastened mattered most.
Screws beat spiral nails. Spiral nails beat common nails. Every time.
Two gussets beat one. Adding a second gusset lifted the same screwed build from 164 lb (730 N) to 298 lb (1,325 N).
One weak joint sinks the whole post. A build with screws everywhere but a spiral-nailed gusset still failed at 112 lb (500 N).

The most important section on this page

You cannot tell by looking

This is the finding that should change how you inspect.

The test crew reported that several guardrails looked sound and well built, and then broke under a light push. A nailed guardrail and a screwed guardrail look identical from three metres (10 ft) away. They are not the same guardrail.

There is a second trap. Nailed guardrails tend to fail slowly. The post leans, the base lifts, and the wood often shows no damage afterward. Screwed guardrails hold much more, then fail all at once, often snapping the post.

A weak guardrail warns you. A strong guardrail does not. Neither one is a reason to lean on it.

So the only reliable inspection is at the fastener, not the eyeball. Get close. Check what is actually in the wood.

Why you must never pull just one rail

A guardrail is a system, not a set of posts.

When someone hits the rail, the rails pull tight and pass part of the load along to the posts on either side. That sharing is why middle posts in the tests held 15% to over 50% more than end posts.

Removing a horizontal rail leaves the posts looking untouched, so it feels harmless. It is not. The load can no longer travel sideways to the next post.

Common site mistake

Pulling the mid rail to pass material through, or knocking a rail off and leaving the posts up. The posts still look like a guardrail. They are now a much weaker one.

Five ways guardrails broke

Learn these and you will know what to look for.

1. The post pulled off the gusset

Nails holding the gusset to the post let go. The post pulled straight off them. The wood was often undamaged, which makes this one easy to miss afterward.

2. The foot split along the grain

Screws holding the gusset drove into the foot and acted like a wedge, opening the wood along the grain. This was the most common failure in the strongest screwed builds.

3. The base plate lifted off the deck

The plate pulled up and out of the deck, taking the whole post with it. This is what happens when the base is nailed to the deck instead of screwed.

4. The post sheared across the grain

The post snapped right at the top of the gusset, where the fasteners are. This was sudden and total.

5. The gusset screws sheared off

The screws themselves broke. In those tests the report noted there may not have been enough screws at that joint.

Watch this one

The toe board fastener. In two tests, the fastener driven through the toe board cracked the base of the post and the post failed early because of it. Keep toe board fasteners out of the bottom of the post.

Build it so it holds

Everything below comes from the test results or current IHSA guidance.

Do this

Screws at all three joints. Gusset to post. Foot to base plate. Base plate to deck. No exceptions, no "just this one".
Use two gussets, not one. The single cheapest strength gain in the whole test set.
Lay the top rail flat, wide face horizontal. IHSA states this gives the best resistance to a sideways push.
Fasten rails so they cannot be pulled off the posts. Fasten to the top of each post, not just the side.
Build to 0.9 m (35 in), not 1.1 m (43 in). The law allows either. Quebec testing found 1 m guardrails held more than 1.2 m (4 ft) ones. A shorter post gives the push less leverage on the base.
Bring posts closer together. The law allows 2.4 m (8 ft). IHSA recommends 1 m to 2 m (3.3 ft to 6.5 ft) for a stronger rail. Doubling the 2x4 top rail helps too.
If nails are all you have, add a metal angle bracket. Quebec testing found a steel angle bracket used with plain nails brought the post-to-deck connection up close to a screwed one. Screws are still better.
Stagger your fasteners. A row of fasteners in a line along the grain creates a split line. Offset them.
Use S-P-F, construction grade or better, with no visible defects.

Never do this

Common nails anywhere in the load path.
Spiral nails as a "good enough" substitute for screws. They tested better than common nails and still failed the 150 lb (675 N) rule at end posts.
Reusing damaged, split, or knotty stock because "it is only a guardrail".
Leaving the posts up after removing a rail.
Sitting, leaning, or stacking material on a guardrail.
Driving a toe board fastener into the base of the post.

Same board, turned two ways. Flat puts the wide face of the 2x4 in the direction of the push, which is where it resists best.

Taking a guardrail down, safely

Landing material is the moment guardrails come off and people fall.

Tie off first. Workers in the unprotected area use travel restraint if possible, fall arrest if not.
Cordon off at least 2 m (6.5 ft) each side of the opening and 2 m (6.5 ft) back from the edge.
Post warning signs at the opening.
Protect the drop zone below. Anything you drop at the edge lands on someone. Before a rail comes off, barricade the area underneath, post signs, and keep everyone out until the work is done. If people must pass, use a spotter to hold the zone. A toe board stops a rolling tool; it does not stop the one in your hand.
Tether your tools. At the edge, every tool gets a lanyard to your wrist, belt, or an anchor - drills, bars, and hammers included. Loose parts (nails, screws, brackets) go in a closed pouch or bucket with a lid, never loose on the deck. A 2 lb (0.9 kg) hammer off a second storey hits like a brick.
Unfasten with proper tools. Do not pry or pull posts off the deck. Prying wrecks the wood and the next crew inherits it.
Store the parts out of the way so nothing gets walked on or lost.
Put it back the same day, by a competent worker, with the same type and number of fasteners as the original.

The rebuild is the risk

The tested guardrails were built once, by a careful crew, in a shop. A guardrail rebuilt in a hurry at the end of a shift, with whatever fasteners are in the pouch, is not the guardrail that was tested. Rebuilds are where the nails creep back in.

Why this still matters in 2026

The tests are from 2003. The problem has not gone away.

16,124 Ontario claims for falls to a lower level, 2019 to 2023. About $318 million paid, 71 lost days per claim on average.

~31% Share of all Ontario workplace traumatic deaths that were in construction in 2023, from 7.5% of the workforce.

63% Of the fall deaths reviewed by Ontario's Chief Coroner, the worker was exposed to the hazard with no fall protection, or was using it wrong.

13 of 26 Fall deaths reviewed where the employer had fewer than 10 workers. Half the workers had been in the job under a year.

Ontario's first coroner-led Construction Death Review, released in 2025, found that most workers who died had completed Working at Heights training. Training was not the gap. What happened on the deck was.

Two things worth knowing from outside Ontario

The US requires more. US federal rules ask a guardrail to hold 200 lb (890 N) sideways. Ontario asks 150 lb (675 N). A rail that just squeaks past in Ontario would fail an American test.
Engineers test to 2.5 times the number. In North American guard research on wood decks and balconies, test loads are set at 2.5 times the required load, because a real guardrail built on site is never as good as one built in a lab. That research found the same thing the Ontario tests found: the connection to the structure fails long before the wood does.

Put those together and the 150 lb (675 N) line is a floor, not a target. If your guardrail averages 164 lb (730 N) at the end posts, you passed with nothing to spare. Two gussets and full screws get you to roughly 300 lb (1,330 N), and that is where you want to be.

What newer research has found since

The 2003 Ontario tests were not the last word. Quebec kept going.

Someone finally tested a real fall, not a slow pull

Every number on the chart above came from a slow, steady pull. A worker tripping into a rail is not slow and not steady. Quebec researchers built a test for that: a 100 kg (220 lb) wooden torso on a rolling frame, hitting the rail at 2 m per second (6.5 ft/s). About walking-into-it speed, with a full-sized body behind it.

The finding that should worry you

The crash test showed that a plain 2x4 post, on its own, is not strong enough to stop a falling worker. The post needs help: plywood gussets at the base, or a second 2x4 beside it.

Ontario law allows the plain post, with no help. So a guardrail can be fully legal and still be the post that failed this test.

What to do about it: gusset every post. Do not build a bare post just because the regulation lets you.

A much bigger Quebec test programme, 2019

More than 350 lab tests on job-built 2x4 guardrails, fastened the way they actually get fastened on a house: to open-web joists and to prefabricated walls. Plus 36 tests on real roofs. What it found:

Screws and lag screws beat nails. Same answer as Ontario got in 2003, on a much bigger sample.
The worst push is straight onto a post, not halfway between two posts. That is the case to build for.
Shorter guardrails held more. 1 m (3.3 ft) beat 1.2 m (4 ft).
A steel angle bracket with plain nails was a real improvement over nails alone.
Deeper joists held better. On prefabricated walls the nailing depth was under an inch, and that was where it let go.

Europe now tests guardrails by dropping weight on them

The European standard for temporary edge protection was rewritten in 2025. Depending on the class, systems must survive a drop-weight impact test, not just a static push. Ontario still asks only for four static point loads. If you buy a manufactured system, ask the supplier which standard it was tested to and get the report.

What we still do not know

Nobody has retested in Ontario since 2003. The organization that ran those tests no longer exists in that form. The Quebec work is the closest thing we have to current Canadian data, and it was done under Quebec's rules, which are stricter.
Wood decks only. The Ontario tests used formply and OSB sheathing. Concrete, steel, and precast are different jobs.
Wood is not consistent. Individual results in the same setup ranged from 120 lb (535 N) to 330 lb (1,470 N). Moisture, knots, and rot all matter and none of them are visible from a distance.
The 2003 report asked for more testing itself and invited other provinces to share data for comparison. That comparison never happened in Ontario.

None of that changes the main finding. Across every study, in two provinces, over 20 years, screwed connections outperformed nailed connections every single time.

Daily guardrail check

Print this page. Walk the edge. Ten seconds per post.

Top rail, mid rail, and toe board all present along the whole run?
Top rail between knuckle and armpit height on an average adult? (0.9 m to 1.1 m)
Toe board tight to the deck and at least a hand's width high? (89 mm)
Posts no more than 8 feet apart? (2.4 m)
Guardrail within a foot of the edge? (300 mm)
Get close: screws or nails at the gusset, the foot, and the base plate?
Any split wood at the base of a post, or around the toe board fastener?
Any post that moves when you push it with a flat hand?
Any rail that was taken off and not put back?
Anything stored against, hung from, or leaning on the guardrail?

Stop work

If you find common nails in a guardrail base, treat the run as not there. Barricade it back, tie off, and rebuild it with screws before anyone works that edge.

What we are NOT saying

We are not saying guardrails fail. A well-built guardrail holds, and the tests prove that too.

We are not telling anyone to ignore the rule. The rule is the law. Build to it.

Our whole point is two lines:

The rule allows one weak build - a bare post with no bracing. Testing found that build can fail in a real fall.
The rule never names the fasteners. In every test, the fasteners decided whether the guardrail held.

So build to the rule, then build past it: screws, two gussets, posts closer together. That guardrail passes the rule and passes the testing.

Where this comes from

Links go to free official copies where one exists. Journal papers open at the publisher and may need library access.

Construction Safety Association of Ontario. Test Report: 2x4 Wood Guardrail Assembly. October 30, 2003. Technical Services and Quality Assurance. Over 90 static load tests on formply and OSB decks. Not published online - contact us and we will point you to it.
Galy, B. and Lan, A. (2019). Resistance des garde-corps en bois fixes sur des structures neuves et comportement sous charges de garde-corps metalliques fixes sur des structures existantes. IRSST Report R-1048. Open access, free to download (French).
Galy, B. and Lan, A. (2020). Strength of Temporary Wooden Guardrails Installed on Open Joists and Prefabricated Walls. Practice Periodical on Structural Design and Construction 25(2).
Galy, B. (2021). In Situ Strength Testing of Temporary Steel Guardrails Installed on Existing Flat-Roof Parapets. Practice Periodical on Structural Design and Construction 26(1).
Lan, A. and Daigle, R. (2009). Development and validation of a method for evaluating temporary wooden guardrails built and installed on construction sites. Safety Science 47(2). Source of the dynamic impact test and the S-P-F post finding.
EN 13374:2025, Temporary edge protection systems: product specification and test methods. European standard, published May 2025. Sold through national standards bodies; not free online.
Ontario Regulation 213/91, Construction Projects, sections 26, 26.1 and 26.3, as currently in force. Free on ontario.ca - always check the current text.
Infrastructure Health and Safety Association. Construction Health and Safety Manual, Chapter 18: Guardrails. IHSA is the successor organization to CSAO.
Office of the Chief Coroner for Ontario. Construction Death Review: Falls from Heights, annual report, June 2025. Free PDF.
Ontario Ministry of Labour, Immigration, Training and Skills Development, construction compliance campaign data, and WSIB claims data 2019 to 2023.
US OSHA 29 CFR 1926.502, fall protection systems criteria, for the 200 lb (890 N) comparison.
Published North American research on wood guard post connections and the 2.5 safety factor used in guard testing.

This guide summarizes and interprets those sources in plain language. It does not replace them. Where a question is close, read the regulation and get advice.