A near miss is an event that could have caused injury or damage but did not, often by luck or timing. Near misses are free lessons: they show where controls are weak before someone gets hurt. A reporting system works when reports are easy to make, never punished, and visibly acted on.
What counts as a near miss
Definitions vary between programs, but the common thread is an unplanned event with the potential for harm. Examples:
- A load swings close to a worker during a lift.
- A tool falls from a scaffold into an area that happened to be empty.
- A gas detector alarms when a worker opens a tank hatch.
- A forklift and a pedestrian meet at a blind corner and stop in time.
- A circuit believed to be locked out turns out to be live during testing.
Some programs also invite reports of unsafe conditions (a missing guardrail, a damaged extension cord) even when no event happened. Whether you call those hazard reports or near misses matters less than making sure both get captured.
Why they matter
Serious injuries rarely come out of nowhere. The same conditions that produce a near miss today can produce an injury tomorrow if nothing changes. The difference between the two outcomes is often a few centimetres or a few seconds. Treating near misses as warnings lets you fix the conditions while the cost of learning is low.
Near-miss reports also tell you how your program is working. A site with lots of reports and fast follow-up usually has people paying attention. A site with none is more likely to have people who see no point in reporting.
Barriers to reporting
Workers often know about near misses that never reach a form. Common reasons:
- Fear of blame or discipline, for themselves or a co-worker.
- Forms that are long, confusing or only available in the office.
- Nothing ever changes after a report, so it feels like wasted effort.
- Safety incentive programs that reward low incident counts, which quietly discourage reporting.
- A belief that "nothing happened, so it doesn't count".
Most of these are about trust. Our guide to psychological safety and hazard reporting goes deeper on that.
Building a system that works
Make it easy
Keep the form short: what happened, where, when, and what could have happened. Offer more than one way to report: paper cards on site, a phone app, a quick conversation with the supervisor who writes it up. Allow anonymous reports if that helps people start.
Make it safe
Say plainly, and show in practice, that reporting a near miss will not lead to discipline. If a report reveals a deliberate, serious rule breach, deal with that separately and carefully, and explain why to the crew.
Respond quickly
Acknowledge each report within a short time, even if the full response takes longer. Fix simple hazards right away. Feed reports with serious potential into your incident investigation process.
Close the loop
Tell the reporter, and the wider crew, what was done. Use toolbox talks to share recent near misses and the changes they led to. Nothing builds reporting faster than seeing a report turn into a fix.
Do not set targets for fewer near misses. A falling count can mean fewer events or fewer reports, and you cannot tell which. Track how quickly reports are closed and what changed because of them instead.
Sorting and acting on reports
Rate each report by potential severity: what is the worst realistic outcome if the circumstances had been slightly different? That rating decides how deep the follow-up goes.
| Potential severity | Typical response |
|---|---|
| Low | Fix the condition, record it, mention at the next crew meeting |
| Moderate | Supervisor review, corrective action with an owner and date |
| High (could have caused serious injury or death) | Formal investigation, same as an actual serious incident |
Review near-miss data periodically with your health and safety committee or representative. Look for repeats by location, task, equipment or time of day. Clusters often point to a design, scheduling or staffing problem that no single report would reveal.
A typical case
In a distribution warehouse, a picker reports that a forklift came around the end of a racking aisle as she stepped out, and both stopped with little room to spare. The supervisor acknowledges the report the same shift and rates it high: the worst realistic outcome is a pedestrian struck by a loaded forklift. The follow-up treats it like an incident. The team finds that the aisle end has no mirror, the pedestrian route crosses the forklift route at that point without markings, and pickers take the shortcut because the marked walkway adds a long walk. The fixes are to move the pedestrian crossing to a visible spot, add a barrier at the blind end, mark the floor, and set a stop-and-sound rule for forklifts at aisle ends. At the next toolbox talk the supervisor explains the report and shows the changes. Over the following weeks two more reports come in from other aisle ends, which is the system working.
What the report card should ask
- What happened, in a sentence or two
- Where and when
- What could have happened if things had been slightly different
- Anything done on the spot to make it safe
- A suggestion for a fix, if the reporter has one
- Name, optional where anonymous reporting is allowed
Measuring the system, not the count
Useful measures include the time from report to acknowledgement, the time to close actions, the share of reports that led to a change, and whether reports come from every crew and shift. A crew that never reports is worth a conversation, not a commendation.
How ehZee can help
ehZee helps employers design reporting systems that people actually use, connect them to investigation and corrective action, and review trends. See safety program development and review.
Common questions
Do near misses have to be reported to the government?
Some serious incidents with no injury, such as certain structural collapses, equipment failures or uncontrolled releases, are reportable in many provinces. Check the reportable events list for your province and include it in your procedure.
Should we reward people for reporting?
Recognition can help, but keep it about the quality of reports and the improvements that follow, not the number of forms. Simple acknowledgement and visible action usually matter more than prizes.
What if the same near miss keeps being reported?
That is a sign the fix did not reach the cause. Treat the repeat as a finding in its own right, look higher on the hierarchy of controls for a solution, and tell the reporters you have noticed the pattern.