Twenty-eight questions about what actually happens at your workplace — not what the policy says. You get a maturity read on nine areas, and one thing to do next.
Regulatory statusA structured self-assessment for your own planning. Not an audit, not a compliance determination, and not connected to any OSHA requirement or program.
Most self-assessments ask whether you have a policy. That is the wrong question: almost
everyone has a policy, and having one predicts very little. These questions ask what
happens — when someone spots a hazard, when an action passes its date, when
the month gets busy.
Every question offers four descriptions rather than a rating. Pick the sentence that
matches, including when the honest answer is the first one. An assessment answered
aspirationally tells you nothing you did not already believe.
Nothing you answer leaves your browser. Your answers and your results are
not transmitted to EHSTEC, not saved on this device, not written into the address bar, and
not included in any analytics — not even as a summary band, because a band is worked
out from what you told us. Refresh the page and it is gone, so if you want
to keep the result, print it or write it down before you leave.
There is no signup, no email, and no part of the result held back. What you see when you
finish is the whole thing.
The questions
What good looks like:
Anyone can raise a hazard without asking permission or fearing blame, and they find out what happened to it.
What good looks like:
Hazards are found deliberately — before work, after change, and from your own history — rather than discovered by someone getting hurt.
What good looks like:
Inspections happen on a schedule you chose and can justify, findings leave with an owner and a date, and repeats are visible.
What good looks like:
Every action has one issue, one named owner and a real date, and someone other than the owner confirms it actually worked.
What good looks like:
People are trained before the work, on the hazards they will actually meet, and you can tell who is current without asking around.
What good looks like:
Incidents and near misses are investigated for cause rather than blame, and what you learn changes something.
What good looks like:
The records you are required to keep exist, are current, and can be found by someone other than the person who wrote them.
What good looks like:
You track a small number of things that tell you whether the programme is working, and you act on what they say.
Your result appears below and updates as you answer. There is nothing to submit.
Where your program stands
Where to focus first
Ordered by where attention pays most — which accounts for both how weak an area
is and how much the rest of the program depends on it. It is not a ranking of how bad
things are.
The one thing to do next
In :
How the scoring works
Published because a self-assessment that hides its arithmetic is asking to be trusted
rather than understood.
Each question scores 0 to 3 in the order the options are shown, least
to most developed.
Each of the nine areas gets a band from its own total: Not
established below 25%, Emerging below 50%, Developing below 80%,
Established at 80% and above. The bar for Established is deliberately high
— answering “usually” to everything in an area is
Developing, because the thing happens but not reliably.
Areas are weighted, but only for prioritising — never for banding.
An area's band is a fact about that area. Leadership and worker participation carry the
most weight when deciding where to focus, because a program fails there first and
everything downstream inherits it.
An area you do not finish is not scored. It is reported as
Not assessed and named. An unanswered question is never counted as zero
— that would invent a finding about a workplace nobody described.
There is no overall score, and that is on purpose. One band across a
whole program is a grade, and a grade ends your thinking exactly where it should
start. Knowing that corrective actions are Emerging while inspections are Established
tells you where Monday goes.
What this is not
Not an audit and not a substitute for one.
Not a compliance determination. Nothing here tells you whether you
meet any OSHA requirement, and a program can score well here and still have a
violation.
Not endorsed, approved or reviewed by OSHA, and not connected to any
OSHA program.
Not a benchmark. You are not being compared with anyone; there is no
peer data behind this and we are not pretending otherwise.
Not a prediction. It makes no claim about your injury rate, your
insurance, or what any software would change.
Only as honest as the answers. The most useful version of this is
answered by several people separately — a manager and a crew member often pick
different sentences for the participation questions, and that gap is itself the
finding.
Free resources, by area
The whole map, not just the areas your answers surface, so it is useful before you start
as well as after you finish.
We do not have a free resource for this yet. Rather than point you at something that
does not fit, the recommendation for this area is a practice — and the gap is on
our list.
The nine areas follow OSHA's Recommended Practices for Safety and Health Programs,
which are recommendations and not a standard — they carry no legal
requirement. Nothing here is endorsed or approved by OSHA, this is not legal
advice, and completing this assessment does not establish compliance with
anything. What applies to you depends on your industry, hazards and jurisdiction. See our
editorial standards for how this content is written and
reviewed.