Fall Protection Checklist: 7 Steps I Run Before a Crew Goes Up

This checklist is for the safety manager or site supervisor who signs off on fall protection before a crew goes up. Not the procurement guy comparing line items. The one standing at the edge, deciding whether the setup is good enough to work from.

I'm a quality and brand compliance manager at a safety equipment distributor. Every fall protection order crosses my desk before it ships—around 400 line items a quarter. In 2024, I rejected roughly 9% of first deliveries for webbing tolerance, stitching deviations, or labeling that didn't match the spec sheet. Paperwork problems are the easy ones. The field problems are the ones that keep me up.

Seven steps. Run them in order. Step one is the one most people skip.

The Checklist

Step 1: Decide Fall Arrest vs. Fall Restraint First

Before you open a catalog, answer one question: are you stopping a fall, or preventing one from happening at all?

That's the difference between fall arrest and fall restraint. Fall arrest means the worker can still fall—you're catching them mid-air. Fall restraint means the system physically stops them from reaching the edge. Similar hardware on paper. Completely different engineering.

It's tempting to think a lanyard is a lanyard. But a fall arrest lanyard has to absorb energy—typically through a deceleration device that keeps forces under 1,800 lbs per OSHA 29 CFR 1926.502. A fall restraint lanyard has no energy absorber because there's nothing to absorb. Use a restraint lanyard in an arrest application and you've built a system that fails at exactly the moment it's needed.

In 2022, a contractor asked me to quote a "simple lanyard setup" for roof work near an HVAC unit. I assumed restraint. He assumed arrest. We caught the mismatch during a routine spec review, not during use. That near-miss changed how I quote anything now—I ask about working geometry before I'll send a price.

The rule I use: if the anchor point can't keep the worker at least 6 feet back from an unprotected edge, you're in arrest territory. Plan accordingly.

Step 2: Walk the Edges and Mark Them Before You Buy Barriers

The site walk is where the checklist earns its keep. You're not looking at the building. You're looking at every place a worker could end up too close to an edge—open shafts, loading docks, roof perimeters, the top of a retaining wall.

I have mixed feelings about split rail fence as edge protection. On one hand, it's cheap, it goes up fast, and a well-built run of split rail can serve as a visual warning line or a restraint barrier for low-hazard zones. On the other hand, I've seen split rail used as a substitute for engineered guardrail, and it isn't. Decorative split rail has no load rating. A leaning worker can put 200+ lbs of lateral force on a fence—more if they're carrying material. Split rail might hold. It might just lean and dump them.

If you use split rail at all, use it as what it is: a warning line or a low-risk restraint. Not as primary guardrail. Not as a fall arrest anchor. Not as something you point to and say "that's taken care of."

Mark every edge separately. Photograph it. Note the gap distances. A plan that says "perimeter protection as required" is a plan that fails inspection.

Step 3: Match the Harness to the Task, Not the Catalog Photo

This is where brand loyalty and lazy specifying both cause problems.

I get asked about the Honeywell Miller H500 fall protection harness a lot—it hits a decent price point and comes from a brand most safety managers already recognize. On paper it's solid. But solid isn't the same as appropriate.

When a contractor asks whether the H500 is the right choice, I ask three things: anchor height, working duration, and rescue plan. If the rescue plan answer is "we'll figure it out," the harness choice doesn't matter—you're not ready. If the answer is solid, the H500 is a reasonable fit for general construction and maintenance work with an anchor at or above shoulder height and a predictable work angle.

A few months back, every cost spreadsheet pointed to the H500 as the answer for a mid-size warehouse job—cheapest per unit, in stock, familiar to the crew. My gut said the anchor geometry wasn't what we assumed. I pushed for photos. The original drawings showed anchor heights two feet higher than the field measurement. Would the H500 still have worked? Probably. Would I want to bet a fall on "probably"? No.

Step 4: Inspect Every Connector, Lanyard, and Anchor Point

This is the step where "it looked fine" gets people hurt. Not because they're careless—because they're fast. Snaphooks that don't fully close. Swivels that have stopped rotating. Lanyards with abrasion you can see but don't want to see, because replacing it means paperwork.

My rule: if I can find a defect on site in under 30 seconds, it should have been caught before the equipment left the shop.

• Snaphooks: check that the gate seats fully and the locking mechanism moves freely. A hook that "mostly closes" is a hook that will open under load.
• Webbing: run it through your hands. You're feeling for cuts, chemical stiffening, and sun damage. UV exposure changes webbing texture before it changes color.
• Stitching: if a stitch line looks uneven, tag it. My tolerance threshold is 1/8" deviation from the spec drawing. Most field inspectors would never notice that. The lab does.

One rejected batch in 2023—1,200 harnesses—cost the importer $18,000 in rework and pushed a customer's launch back two weeks. The defect was webbing that measured 1.75" instead of the specified 1.73". A 0.02" difference. That's how the spec works.

I'm not going to pretend this is a glamorous part of the job. It's the part where you look stupid for being careful. Do it anyway.

Step 5: Cover Head Protection—Style Is Fine, Certification Is Not

Fall protection is head-to-toe. If the harness is right but the head protection is wrong, you still have a problem.

I see a lot of cowboy hard hats on outdoor crews—ranch work, agricultural sites, and some utility jobs where the wide brim actually helps with sun. I get the appeal. They're comfortable, they shade your eyes, and guys who work outside all day are going to gravitate toward them.

Just confirm the ANSI Z89.1 rating before it goes on someone's head. Cowboy style doesn't exempt you from Type I or Type II requirements. If the brim looks good but the certification is missing, it's a costume, not PPE.

Step 6: Verify Training Records, Not Training Claims

Anyone can say their crew is trained. Very few can produce the records in under ten minutes.

Miller fall protection training is one of the programs I actually trust, partly because the documentation trail is solid. But the brand of the training matters less than these questions: Who delivered it? When? What did it cover? Did it include hands-on component inspection, or was it a video and a quiz?

I ran a blind check on our own training records in Q1 2024, and 18% of the cards were missing the instructor's sign-off. Eighteen percent. That's not a Miller problem—that's us. We fixed the process. But it's a reminder that "trained" is a record, not a reputation.

Step 7: Document the Setup Before the Shift Starts

Last step. Also the one most likely to be skipped when the schedule runs tight.

I had two hours to review a fall protection plan before a crew arrived in early 2024. Normally I'd want a full pre-job meeting and a physical walkthrough. There was no time. I signed off on photos and a written plan. In hindsight, I should have pushed the start time back. I didn't, because the client was already on site and the delay would have cost them a day. Nothing went wrong that day. That's not the same as saying it was fine.

What I want to see in the document:

• Photos of each anchor point with the anchor's rated capacity visible
• The fall arrest vs. fall restraint classification for each work area
• Inspection records for every harness, lanyard, and connector used that day
• A rescue plan—not just "call 911," but the actual steps a trained person takes

The rescue plan is the one that gets left blank. It shouldn't be. OSHA 1926.502 sets a maximum suspension time of 15 minutes before rescue begins. That's not a lot of margin. If your plan doesn't identify who does what in that window, you don't have a plan.

Common Mistakes (and One Honest Limit)

These are the issues I see most often across the 400-SKU review cycle and the site visits in between:

• Treating fall arrest and fall restraint as interchangeable. They aren't. The hardware and the planning are different.
• Using decorative barriers as primary protection. Split rail is fine as a warning line. It's not an engineered guardrail.
• Skipping the rescue plan. The most common blank field on the most common template.
• Confusing brand recognition with task fit. The H500 is a good harness for the right job. It's not a universal answer.
• Assuming training happened because someone said it did. Check the records.

And one thing about my own limits: we don't do engineered anchor design. We don't do rescue consulting. When a contractor asks whether we can design a custom anchor system for their building, I tell them no—and I give them the names of two firms that do it well. We lose that revenue. We gain the next harness order, and the one after that, because the contractor knows we're not going to pretend to be something we're not.

That's the checklist. Seven steps, run in order, checked against the spec. Most of them are boring. The boring ones are the ones that keep people alive.