Job hazard analysis, toolbox talks, and pre-task planning field guide

What the daily safety routine is

The daily safety routine is three tools that put hazard control where the work actually happens. A job hazard analysis, the JHA, breaks a task into its steps and matches each step to the hazard it carries and the control that handles it. A toolbox talk is the short crew briefing that shares one hazard and the plan for it. A pre-task plan is the crew looking at today's specific work and conditions before they start. Run together, they move safety off the shelf and onto the deck.

These are not the written program. The written program is the binder, the policies, the training records, the inspections, and the people behind them, and it lives in the construction safety program guide. This is the part that runs every morning at the truck and at the work face. The program decides what your company does about each hazard in general. The JHA, the talk, and the pre-task decide what this crew does about the hazard in front of them right now.

The reason all three exist instead of one is that a hazard has three problems. It has to be analyzed before anyone does the task. It has to be shared with the crew that faces it. And it has to be checked against the way the site actually looks today, which is never quite the way it looked yesterday. One tool does not cover all three. Skip any of them and you have a gap exactly where the work meets the danger.

Why the five minutes before the task prevents the injury

Most incidents trace back to a hazard that was never caught before the task started. Not a freak event, not bad luck, but a step nobody looked at and a condition nobody flagged. The safety manual sitting in the trailer cannot catch that hazard, because the manual was written for every job and this is the one job in front of you. The five minutes the crew spends looking at the task before they touch it is the only point in the day where the hazard is still cheap to fix.

This is what people mean by the point of work. The program lives or dies there, not in the office. A perfect written plan that the crew never connects to today's task does nothing for the worker on the edge. A rough plan that the foreman actually walks with the crew before the tear-off starts changes how the next two hours go. The closer the safety thinking sits to the moment of the work, the more likely it is to catch the thing that hurts someone.

OSHA expects it, in substance if not always by name. The General Duty Clause requires an employer to keep the workplace free of recognized hazards likely to cause death or serious harm, and you cannot keep a hazard out if you never looked for it. The construction provisions put the duty on the employer to instruct workers in the hazards they will face and to inspect the work as it goes. The JHA, the toolbox talk, and the pre-task plan are how that duty gets met on a real morning, and they are what a general contractor and an insurer look for when they ask how you run safety.

The three tools and how they fit

The three tools answer three different questions, and they hand off to each other. The JHA answers what the hazards of this task are and how we control each one. It is the analysis, usually built once for a task and revisited when the work changes. The toolbox talk answers what the crew needs to hear today and gives them a chance to talk back. The pre-task plan answers what is different about this morning, on this part of the building, with this crew and this weather.

Think of them as the same job at three altitudes. The JHA is the deep look at the task itself, done with enough time to think it through. The toolbox talk is the broadcast, short and focused, where the foreman puts one hazard in front of everyone and the crew adds what they know. The pre-task plan is the close-in check at the work area, where the plan meets the actual conditions and gets adjusted before a hand goes on a tool.

They reinforce each other when they run together and turn into three separate paperwork chores when they do not. The JHA feeds the toolbox talk, because the steps and hazards you wrote down are the topics worth talking about. The talk feeds the pre-task, because the hazard you just discussed is the one the crew now checks for under the conditions on the ground today. The pre-task feeds the next JHA, because what changed today is what you missed last time. The site safety program guide covers the structure that holds all three together; this guide is the routine that runs inside it.

The job hazard analysis: step, hazard, control

A job hazard analysis takes a task apart before anyone does it and asks, at each step, where it can hurt someone and what stops it. OSHA's own guidance, Publication 3071, lays out the method plainly: break the job into steps, identify the hazard at each step, then determine the control. The same process goes by job safety analysis, JSA, and the construction world's activity hazard analysis, the AHA, that many general contractors require for each phase of work. Different names, one logic.

Break the task into the right number of steps, which for most jobs lands somewhere around eight to twelve. Too few and a whole hazard hides inside one vague step like prepare the roof. Too many and you are writing down every hand motion and nobody reads it. The test is whether each step is a distinct piece of work with its own hazard. Set up the ladder is a step. Tear off the old membrane is a step. Lower the debris is a step. Each one carries a different way to get hurt.

The analysis is only as good as the controls in the third column. A step with a hazard and no real control is a warning, not a plan. The control has to actually remove or reduce the hazard, and it has to be something the crew can do on this task with the gear on this truck. A JHA that names every hazard and then writes be careful in the control column is a checkbox, and the section on quality covers why that is the most common way these fail.

The three columns, on a real task

The whole JHA fits in three columns: the step, the hazard at that step, and the control for that hazard. Kept specific to the actual task, it reads like a plan a crew can follow. Below is a roofing tear-off and re-cover broken down the way it would run on the deck, with the hazard and control each step carries.

Notice that the controls are not all PPE. Some steps eliminate the hazard, some engineer it out, some are procedure, and gloves show up only where they are the right answer. That mix is the point of the next two sections.

Build the JHA with the crew who do the work

A JHA written at a desk by someone who has not done the task is a guess. The people who run the work know the step where the footing is bad, the spot where the line has to come unclipped to move, the corner where the wind comes over the parapet and grabs a sheet. Those are the steps that hurt people, and they are exactly the ones a desk version smooths over. Build the analysis with the crew, and walk the task with them where you can.

Walking the task beats imagining it. Stand at the access point and the step about the ladder writes itself. Stand at the edge and the fall hazard stops being abstract. Walking it also surfaces the steps a written list skips, the move between work areas, the material handling, the cleanup, which is often where the trip and the strain live. The construction safety program guide makes the same point about the broader program: plan it with the people on the roof, not above them.

For each hazard, work down the hierarchy of controls in order rather than reaching for the same answer every line. Can you eliminate the hazard, design the exposure out of the task entirely? If not, can you substitute a safer method, engineer a barrier, or set a procedure that keeps the worker clear? Only when those run out does the control become PPE. Building the JHA this way, hazard by hazard, is what separates a real analysis from a form that lists a hard hat against every line.

Apply the hierarchy of controls, not PPE on every line

The hierarchy of controls ranks the ways to handle a hazard from most effective to least: eliminate it, substitute something safer, engineer it out with a guard or barrier, control it with procedure and training, and only then fall back to personal protective equipment. The order is not a preference. It is a ranking by how likely the control is to fail. Eliminating a hazard removes it for everyone, every time. PPE depends on one person wearing the right gear correctly on every exposure, which is the line most likely to break.

The most common JHA failure is a third column that reads PPE on every step. Wear gloves, wear a harness, wear glasses, all the way down. PPE has its place, and on a roof the harness is often the honest answer at the edge. But when every control is PPE, the analysis has skipped the question that prevents the injury: can we keep the worker away from the hazard in the first place? A guardrail that removes the fall exposure protects the worker who forgot to clip in. A harness only protects the worker who remembered.

Use the hierarchy as the prompt for each hazard, in order, and the better controls show up. The leading edge: can the work be done back from it, can a guardrail go up, before the answer becomes arrest gear. The cut from the saw: can a different tool or a different sequence reduce the exposure before the answer is a better glove. The fall protection treatment in the construction safety program guide runs the same logic for heights, remove the exposure first and protect against it last, and a JHA that honors that order is doing its job.

The toolbox talk: short, relevant, two-way

A toolbox talk is the short safety briefing that puts one hazard in front of the crew before the work starts. Five to fifteen minutes, run at the truck or the gangbox, on a topic that matches the day. It goes by other names, tailgate talk, safety moment, pre-job briefing, and the labels do not matter. The format is what matters: short enough that the crew stays with it, focused enough to land one point, and open enough that the crew talks back.

Keep the cadence regular. Many crews run a talk daily before the work, others weekly with a quick daily huddle on top, and many general contractors and state plans require a documented frequency on their sites. The construction safety program guide covers where that cadence fits in the overall program. The frequency matters less than the honesty: a talk run because it is on the schedule, read off a generic page word for word, does nothing. A talk that names the actual exposure on this roof, this morning, changes the next hour.

Two-way is the part most crews get wrong. A foreman reading a printed page at the crew is a lecture, and a lecture rolls off. Ask the crew where the hazard shows up on this job, ask the worker who almost got hurt last week what he saw, and the talk turns into the crew's own knowledge instead of a handout. The engagement section goes deeper on this, because a talk nobody participates in is a talk nobody remembers. Document the topic, the date, who led it, and who attended, every time, because the sign-in is the proof the crew was briefed on the exact hazard.

Good toolbox talk topics come from the day

The best topic is the hazard the crew is about to face. If the morning is tear-off at the leading edge, the talk is fall protection at the edge, not a generic page on office ergonomics that happened to be next in the binder. The topic should be close enough to the work that a worker can walk straight from the talk to the task and use what he just heard. A talk disconnected from the day is a chore everyone waits out.

A recent near-miss is the strongest topic there is, because it actually happened to this crew. The line that was supposed to be dead, the sheet the wind almost took, the ladder that shifted, these carry weight that a printed scenario never will. Pull the near-miss from your own reports, name it without naming blame, and the crew leans in because it could have been any of them. The continuous-improvement section covers how those near-misses feed back into the talks and the JHAs.

Season and conditions drive the rest. Heat as the forecast climbs, hydration and the signs of heat illness when the deck turns into an oven. Cold and ice when the surface freezes overnight, which the jobsite housekeeping guide treats in detail for traction. Wind on a day a roof is open and material wants to fly. Match the talk to what the day is throwing at the crew, and the talk earns its five minutes instead of burning them.

The pre-task plan: the crew looks at today

A pre-task plan is the crew's field-level look at the actual work and the actual conditions, right before they start. It goes by pre-task planning, the PTP, or field-level risk assessment, the FLRA, and it is the tool that closes the gap between the analysis and the day. The JHA was built for the task in general. The pre-task plan asks how that task lines up with this morning: the weather on the deck, the trades sharing the area, the equipment that showed up, the people on the crew, the part of the building you are working today.

It is a living check, not a static document. A pre-task plan filled out once and copied for the rest of the job is a JHA with a different cover. The whole value is that it is redone for today, because today is what changes. The crew stands at the work area, runs through the steps and hazards, names what is different from yesterday, and decides who does what before a tool comes out. Five minutes, at the point of work, with the people who are about to do it.

The pre-task is also where stop-work authority gets real. The crew that just talked through the hazards and named the controls is the crew that will notice when one of those controls is missing, and the pre-task is the moment to set the expectation that anyone can stop the work if it goes sideways. The stop-work section ties the two together. A pre-task plan that ends with the crew clear on the hazards and clear on their authority to stop is worth more than any form they signed.

What changed today is the whole point

The reason a pre-task plan happens every day is that the hazard that hurts you is usually the one that is different today. Yesterday the deck was dry and the south side was clear. Today it rained overnight, a sheet-metal crew is working the same elevation, and the hoist is staged where the walkway used to be. The task is the same on paper. The conditions are not, and the new condition is the one nobody planned for, because the plan was built for yesterday.

Run through the short list of what changes and you catch most of it. Weather, first, because it changes traction, wind, heat, and visibility faster than anything else. New trades in the area, because a crew working below or beside you creates struck-by and access hazards that were not there when you had the space to yourself. A new work area, with its own edges, openings, and footing. New equipment, a different lift, a borrowed saw, a hoist set up overnight. And new people, the worker who is on this task for the first time and does not know the spot where the footing is bad.

This is the dynamic hazard, and it is the one a static binder cannot reach. The written program is fixed; the jobsite moves every day. The pre-task plan is the only tool in the routine built to catch the change, which is why a crew that skips it on a day that looks routine is exactly the crew that walks into the thing that changed while they were not looking. Ask the question out loud every morning: what is different today, and what does it do to our plan.

SLAM and take-5: the habit at the task

SLAM and take-5 are the individual version of the pre-task plan, done by one worker at the moment of the task. SLAM stands for stop, look, assess, manage: stop before you start the step, look at what is around you, assess the hazard, and manage it before you go. Take-5 is the same idea, a sixty-second pause to check the work area, the conditions, and your own readiness before the first move. It is a last-minute risk assessment, and it sits inside the crew's pre-task, not instead of it.

The reason it matters is that conditions change between the morning plan and the afternoon task. The crew planned the day at the truck, but six hours later the worker is on a different part of the roof, the light has changed, the surface is wetter, and the original plan did not see this exact moment. The take-5 is the worker's own check that the plan still fits what is in front of him right now. It is the habit that catches the change the crew briefing was too early to see.

Built into a crew, it becomes a reflex rather than a form. The worker who pauses at the edge, looks for the anchor, and checks his connection before he leans out is running a take-5 whether he calls it that or not. Teach it as a habit and the crew carries it everywhere, on every step, on the days nobody is watching, which is exactly where the injuries happen.

Make it two-way or it does nothing

The single thing that separates a safety routine that works from one that does not is whether the crew owns it. A foreman lecturing a crew that stands there waiting for it to end is going through the motions, and the crew knows it, and the hazard goes uncaught because nobody was actually thinking about it. The crew has to find the hazards, name them, and decide the controls, because the knowledge that prevents the injury is in their hands, not on the page.

Two-way is a skill, and it starts with asking instead of telling. Where does this task bite, on this roof? What almost got someone last week? What is different about the south side today? The worker who has done this tear-off a hundred times has seen things the JHA never captured, and the only way that knowledge gets onto the plan is if someone asks for it and then listens. A crew that contributes to the plan defends the plan. A crew that gets handed the plan ignores it the moment the foreman turns around.

Ownership is the culture this builds, and it is slow to grow and fast to lose. It comes from the crew seeing that their input changes the plan, that the hazard they raised got fixed, that the stop-work call they made got backed. Treat the talk and the pre-task as a form to sign and you train the crew to sign forms. Treat them as the crew's own look at their own work and you build a crew that watches out for the hazard whether or not anyone is making them.

Stop-work authority belongs in the pre-task

Stop-work authority is the explicit, real right of any worker to stop a job when they see a hazard, with no fear of being blamed for slowing the work. It is not a poster. It is a standing rule that the crew believes, and the only way they believe it is by watching it get used and backed. The pre-task plan is where it gets set every day, because the crew that just talked through the hazards is the crew most likely to notice when a control is missing and most ready to call it.

The hard part is the no-blame, and it is tested the first time a stop-work call turns out to be a false alarm. The worker stopped the lift because something looked wrong, and it was fine. If that worker gets grief for it, stop-work authority is dead on that crew, and the next time something actually is wrong, nobody says a word. Back the call even when the worker was wrong, because the cost of one paused task is nothing against the cost of the call that never came. The construction safety program guide treats stop-work and no-blame reporting as the heart of the culture, and the pre-task is where that culture meets the day.

Tie it to the pre-task out loud. End the morning plan with it: if any control we just named is not in place, or you see something we did not plan for, stop the work and we sort it out. That one sentence turns stop-work from a policy in the binder into a permission the crew actually has. The worker who hears it every morning is the worker who uses it on the morning it matters.

Document the JHA, the talk, and the pre-task

Everything in this routine produces a record, and the record is what proves the routine is real. The JHA with the steps, hazards, and controls. The toolbox talk with its topic, date, and attendance. The pre-task plan with today's conditions and the crew on it. The near-miss that got reported and fed back. To OSHA, to the general contractor, and to your insurer, the talk nobody signed and the plan nobody wrote did not happen, no matter how good the conversation was.

Attendance and signatures are the part that gets skipped and the part that gets asked for. After an incident, the first question is whether the crew was briefed on the hazard involved, and the toolbox-talk sign-in is the answer. The JHA signed by the crew that built it shows the work was planned before it started. The pre-task plan, dated to the day with the conditions noted, shows the crew checked today and not just copied yesterday. These records are the OSHA and insurance proof that the routine ran, and they are exactly what a compliance officer reaches for.

Paper loses this fight. The sign-in sheet rides around in a truck and disappears, the JHA stays in a binder in the trailer, the pre-task form gets rained on and tossed. A field tool like FieldOS lets a foreman run the talk, capture the sign-in with a tap, build the JHA and the pre-task from a phone at the work area, and tie each one to the job and the date, so the record is created at the moment the work happens instead of reconstructed under pressure when someone official is in the trailer asking to see two years of talks. The point is not the software. The point is that the proof exists because it was made on the spot, by the person who did the work.

Redo it when the work or conditions change

A JHA is not once-and-done. It is built for a task, and the moment the task changes, the analysis is stale. New equipment, a new method, a new material, a new sequence, a new part of the building, any of these can introduce a hazard the original JHA never saw. The same goes for the pre-task: the plan is for today, and a real change during the day, a storm rolling in, a trade moving into the space, a piece of equipment breaking down, means the plan gets revisited before the work continues.

The trap is treating the analysis as a permanent document because redoing it feels like overhead. It is not overhead. It is the difference between a plan that matches the work and a plan that matches some earlier version of the work that no longer exists. The crew that upsized from a guardrail to arrest gear because the edge condition changed, but never updated the plan or the talk, is running on a plan that lies about what they are doing.

Set the trigger explicitly: when the task changes or the conditions change, the JHA and the pre-task get a fresh look before the work goes on. Not a full rebuild every time, but a real check of whether the new condition broke a control or added a hazard. The dynamic hazard is the one that changed; the routine only catches it if the routine itself is allowed to change with the work.

Coordinate across trades on a multi-employer site

A JHA that only sees your own crew is incomplete on a site where other trades share the space. A roofing crew working above masons, a torch running near another trade's combustibles, a hoist swinging where electricians are pulling, each creates an exposure that crosses the line between employers. The hazard one trade introduces lands on the crew next to it, and the analysis has to account for the trades that are actually in the area, not just the task in isolation.

Under OSHA's multi-employer worksite doctrine, more than one employer can be cited for a single hazardous condition, the one that created it, the one that controlled the area, and the one whose workers were exposed. That is the legal reason to coordinate. The practical reason is that the hazard you did not know about, from the trade you never talked to, is the one that gets your crew. Coordinate the JHAs and the pre-task plans across the trades sharing the area, through the general contractor's planning where the contract sets it up, so the overlapping hazards get named before two crews collide.

The simplest version is the crew knowing who is working above, below, and beside them today, and what those trades are doing. That is a pre-task question on a shared site: who else is in our area, and what hazard do they bring. The exact coordination requirements are set by the general contractor, the site rules, and the applicable standards, so confirm what your site demands. The subcontractor safety and orientation pieces sit in the construction safety program guide; the daily coordination is a line on the pre-task.

Tie the JHA to the focus four and the trade's top hazards

When you build a JHA, the hazards to look hardest for are the ones that actually kill people. OSHA's Focus Four, falls, struck-by, caught-in or between, and electrocution, account for the large majority of construction deaths, and falls lead by a wide margin and are the top killer in roofing. A JHA that breezes past the fall step to spend three lines on a paper cut has its priorities backward. Aim the analysis where the bodies are first, then handle the rest.

For roofing the leading hazard set is concrete: falls from the edge, through openings, and through skylights, which the construction safety program guide and the fall-protection material treat in depth. Struck-by from debris coming off the roof and from material handling. Burns and fire from torch and hot-air work. Heat on an exposed deck. The JHA for a given task should call out the focus-four exposures it touches and put the strongest controls against them, because those are the steps with the worst outcome if the control fails.

Naming the hazard category on the JHA also makes the talk and the pre-task sharper. If the task's big exposure is a fall, the toolbox talk is fall protection and the pre-task checks the anchors and the edge today. If it is struck-by, the talk is the drop zone and the pre-task checks who is working below. Matching the daily tools to the focus-four hazard the task carries is how the routine stays pointed at what matters instead of spreading thin across everything.

Quality, not a checkbox: the number-one failure

The most common way these tools fail is the checkbox version. A JHA copied from another project, generic to no task on this site, with another job's hazards still in it, signed by a crew that never read it. A toolbox talk read off a page that has nothing to do with today's work. A pre-task plan photocopied for the week. The form exists, the signature is on it, and not one of them did anything to catch a hazard, because going through the motions is not the same as doing the work.

A checkbox JHA is worse than none, because it shows intent without practice and it teaches the crew that safety is paperwork. After an incident, a generic JHA that obviously does not match what the crew was doing is a finding, not a defense. A thorough, task-specific JHA that matches what was actually happening on site shows your team was managing the risk in front of them. The inspector can tell the difference in about ten seconds, and so can the crew, every single morning.

Making it real is not complicated, it is just work. Build the JHA for this task with the people who do it. Run the talk on today's hazard and let the crew talk. Redo the pre-task for today's conditions. Sign it because you read it, not to fill a box. The whole routine is only as honest as the worst-run piece of it, and the checkbox is the failure that quietly turns a safety program into a stack of paper that proves nothing when it is asked to.

The foreman leads the pre-task and sets the tone

The daily routine rises and falls on the foreman. The crew reads the real standard off what the lead does, not what the manual says. A foreman who runs the toolbox talk like it matters, asks the crew for the hazards, and walks the pre-task at the work area trains a crew that takes it seriously. A foreman who mumbles through a generic page and signs the form trains a crew that does the same. The tone of the whole routine is set by the person leading it.

Leading the pre-task is the daily habit that holds it together. The foreman stands the crew at the work area, runs the steps, asks what changed today, names who does what, and confirms the controls are in place before the work starts. It is five minutes, and it is the most important five minutes of the foreman's day, because it is the only point where the whole crew is thinking about the hazard at the same time before anyone is exposed to it.

Setting the tone goes past the talk. The foreman who clips in on the same roof he tells the crew to clip in on, who stops the job for the hazard, who backs the worker that made a stop-work call, makes the routine real in a way no form can. The construction safety program guide makes the same point about leadership across the program: lead it from the front or do not expect it from the crew. The pre-task is where that leadership shows up every morning.

Feed near-misses back into the JHAs and talks

A near-miss is a free lesson, and the routine only learns from it if the lesson goes back into the tools. The sheet the wind almost took, the line that was dead when someone touched it, the ladder that shifted, each one is a hazard that announced itself without hurting anyone. Capture it, find the cause, and feed it into the JHA for that task and the next toolbox talk, so the thing that almost happened once does not get its second chance.

This is how the JHA stops being a static document and becomes a record of what your crews actually learned. When a near-miss shows that a step had a hazard the analysis missed, the analysis gets updated. When an incident shows that a control failed, the control gets changed. The same near-miss makes the strongest toolbox topic there is, because it happened here, to this crew, on this kind of work. The construction safety program guide covers no-blame reporting and incident investigation as the engine behind this; the daily tools are where the lesson lands.

The loop has to close to be worth anything. A near-miss reported and then filed and forgotten taught nobody. A near-miss reported, traced to a cause, written into the JHA, and raised in the next talk changes how the next crew runs the task. Reward the report, fix the hazard, and keep the blame for willful violations, not for honesty, or the reports dry up and the routine goes blind.

The system: the JHA library, the talk log, the pre-task records

Over time the routine builds a library, and the library is an asset if you keep it and a liability if you lose it. The JHAs for your recurring tasks, refined every time a near-miss teaches you something. The log of every toolbox talk with its attendance. The pre-task records that show the crew checked the conditions each day. Together they are the proof that the routine ran, and the starting point for the next job instead of a blank page every morning.

A JHA library means you are not rebuilding the analysis for a standard tear-off from scratch each time. You pull the one you have refined, adjust it for this task and this building, and the crew starts from a plan that already learned from the last ten jobs. The talk log shows the cadence held and the topics matched the work. The pre-task records show the routine touched the day, not just the binder. The construction safety program guide and the jobsite housekeeping guide cover the inspection and housekeeping records that sit alongside these.

Keeping the system on paper is how it gets lost. A digital system holds it because the JHA, the talk, and the pre-task get built and signed on a phone at the work area and stay tied to the job and the date in one place. A foreman pulls last week's JHA, updates it for today, runs the talk, and captures the pre-task, and the whole library is searchable when the general contractor or OSHA asks for it. The record gets made because making it is part of doing the work, not a second job after the crew has gone home.

Where the daily routine breaks down

The routine fails the same handful of ways on every kind of job, and none of them are exotic. The checkbox JHA, copied and generic, signed without being read, is the most common. The generic toolbox talk, read off a page that has nothing to do with today, runs a close second. Then the pre-task that never happens on the day that looked routine, which is the day the thing that changed walks in unnoticed.

Underneath those sit two failures that quietly break the whole thing. Controls that are all PPE, because the analysis never asked whether the hazard could be removed before it asked the worker to wear something. And no documentation, so the talk and the plan that did happen cannot be proven when an inspector or an insurer asks. A routine that ran but left no record cannot defend the company, and a routine that left a record but never really ran cannot protect the crew.

The last and worst failure is no real stop-work authority. The crew that has no permission to stop, or that watched a stop-work call get punished once, is the crew that keeps quiet when something is wrong. Every other piece of the routine can be in place, and that one missing piece means the hazard the crew saw and did not stop is the one that gets someone. Fix these six and the routine does what it is for.

What to document

Each piece of the routine produces a record, and each record answers a different question after the fact. Capture enough that the next person, or the inspector, understands what the crew planned, who was briefed, and what they checked. The table below is the short list of what to keep and why it earns its place.

Common mistakes

• A copied checkbox JHA nobody reads, generic to no task actually on the site.
• A generic toolbox talk read word for word, unrelated to today's work.
• No pre-task plan for what changed today, so the new hazard is the one that bites.
• Listing PPE as the only control on every line of the JHA.
• No documentation or attendance, so the talk and the plan cannot be proven.
• No real stop-work authority, so the worker who saw the hazard kept quiet.

Field checklist

Standards and references

OSHA does not impose one blanket rule that names the JHA, the toolbox talk, and the pre-task plan and requires them by those words for every job. What it does is set the duties these tools satisfy. The General Duty Clause, Section 5(a)(1) of the OSH Act, requires the employer to keep the workplace free of recognized hazards likely to cause death or serious harm, which you cannot do without assessing the hazards before the work. The construction general provisions at 1926.20 put the duty on the employer to initiate and maintain safety programs and to inspect the work, and 1926.21 requires instructing workers in the hazards they will face.

The job hazard analysis method itself is laid out in OSHA Publication 3071, which describes breaking the job into steps, identifying the hazard at each step, and determining the control, with the controls chosen by the hierarchy of controls. PPE hazard-assessment requirements at 1926 Subpart E and the general-industry counterpart at 1910.132 reinforce the duty to assess before you protect. Many general contractors, owners, and state-plan jurisdictions require documented JHAs or activity hazard analyses, toolbox talks on a set cadence, and daily pre-task planning by contract or site rule, which often go beyond the federal floor.

The exact section numbers, thresholds, and requirements shift between code cycles and between federal OSHA and the state plans, and a state plan can be stricter than the federal standard. The company safety program, the general contractor's site rules, and the authority having jurisdiction control what you actually have to do, so confirm the adopted standards and the contract before you rely on any single citation here. Throughout, three things hold up no matter the jurisdiction: break the task into step, hazard, and control with the hierarchy applied to each; let the pre-task catch what changed today; and make the routine real, two-way, and documented.

Terms and acronyms

The daily safety routine carries a stack of acronyms, and the same idea often shows up under more than one name across OSHA text, the general contractor's paperwork, and the crew's habit. Knowing the terms is part of speaking the language a prequalification form and an inspection both use.

JHA / JSA / AHA

Job hazard analysis, job safety analysis, or activity hazard analysis; the same process of breaking a task into step, hazard, and control

Toolbox talk

A short crew safety briefing, often 5 to 15 minutes, on a relevant hazard; also called a tailgate talk or pre-job briefing

Pre-task plan / FLRA

The crew's field-level risk assessment of today's specific work and conditions, done before the task starts

Hierarchy of controls

The order for handling a hazard: eliminate, substitute, engineer, administrative controls, and PPE last

SLAM / take-5

The individual last-minute risk check at the task: stop, look, assess, manage, in about a minute

Stop-work authority

Every worker's explicit right to stop unsafe work with no blame for the call

Near-miss

An event that could have caused injury but did not; a free lesson that feeds back into the JHA and the talk

Related tools

Calculators and readiness checks for this work

Compare your options

FAQ

People also ask

• What is a construction safety program?
• Why is housekeeping important on a jobsite?
• What is a PPE hazard assessment?
• What is roof asset management?
• How do you investigate a workplace accident?