Hazard control is the process of eliminating or reducing workplace dangers so they can’t injure, sicken, or kill workers. It’s the core of occupational safety: once you’ve identified a hazard, hazard control is how you actually do something about it. The framework most safety professionals use organizes control options into five levels, ranked from most effective to least effective, known as the hierarchy of controls.
The Hierarchy of Controls
The hierarchy of controls is a five-tier ranking system developed by the National Institute for Occupational Safety and Health (NIOSH). It gives workplaces a structured way to choose the best possible solution for any given hazard. The five levels, from strongest to weakest, are:
- Elimination — Remove the hazard entirely
- Substitution — Replace the hazard with something less dangerous
- Engineering controls — Physically isolate workers from the hazard
- Administrative controls — Change how or when people work
- Personal protective equipment (PPE) — Protect the individual worker’s body
The idea is straightforward: always start at the top. If you can eliminate the hazard completely, nothing else is needed. If you can’t, move down to substitution, then engineering controls, and so on. In practice, most workplaces use a combination of levels. A manufacturing plant might substitute a less toxic chemical (substitution), install ventilation hoods over workstations (engineering), limit how long each worker spends in the area (administrative), and require respirators as a backup (PPE).
Elimination and Substitution
Elimination is the gold standard. If the hazard doesn’t exist, no one can be harmed by it. Real-world examples include doing work at ground level instead of at heights, stopping the use of noisy processes, redesigning warehouse intersections to remove blind spots, and automating material delivery with conveyors so workers don’t share space with forklifts. One OSHA recommendation for a busy warehouse: restrict foot traffic so that only people essential to the operation pass through areas where fork trucks are running.
When elimination isn’t possible, substitution is the next best option. This means swapping a dangerous material, tool, or process for a safer one. A print shop might switch from solvent-based inks to plant-based alternatives. A warehouse might replace powered forklifts with walk-beside models or hand trucks, cutting the risk of high-speed collisions. A factory could switch to a process that uses less force, lower temperatures, or reduced electrical current. Like elimination, substitution addresses the root cause of the danger rather than relying on workers to protect themselves.
Both of these top-tier controls share an important advantage: once they’re in place, they require almost no ongoing effort. You don’t need to train people to use them correctly, and they don’t depend on anyone remembering to follow a rule every single shift.
Engineering Controls
Engineering controls physically separate the worker from the hazard. They don’t remove the danger, but they put something between it and the people nearby. Common examples include machine guards that prevent hands from reaching moving parts, ventilation systems that pull chemical fumes away from breathing zones, sound-dampening enclosures around noisy equipment, and safety interlocks that shut a machine down if a guard is opened.
Engineering controls rank below elimination and substitution because the hazard still exists. The ventilation system can break down, the machine guard can be removed. But they’re significantly more reliable than the two lower tiers because they don’t depend on human behavior. A properly installed guard works whether the operator slept well last night or not.
Administrative Controls
Administrative controls change the rules around work rather than the physical environment. They include things like rotating workers through a noisy area so no one person gets a full shift of exposure, scheduling hazardous tasks during low-traffic periods, posting warning signs, writing standard operating procedures, and requiring training before someone operates certain equipment.
These controls are cheaper and easier to implement than engineering solutions, which is why many workplaces lean on them heavily. But they have a significant weakness: they depend entirely on people following the rules, every time, without fail. The Health and Safety Executive, the UK’s workplace safety regulator, warns against assuming that workers will always follow procedures. People forget steps, take shortcuts under time pressure, and sometimes deliberately skip rules they see as unnecessary. Training alone doesn’t reliably prevent the kind of slips and lapses that happen when someone is tired, distracted, or rushing. Any hazard control plan that relies mostly on administrative controls is only as strong as the least attentive person on the shift.
Personal Protective Equipment
PPE sits at the bottom of the hierarchy for good reason. Hard hats, gloves, safety glasses, respirators, earplugs, and fall harnesses all share the same fundamental limitation: they do nothing to reduce the hazard itself. They only protect the individual wearing them, and only if the equipment fits properly, is worn correctly, and is used every single time.
A respirator that slips, a pair of gloves with a tear, earplugs that aren’t inserted deep enough: any of these failures puts the worker right back at full exposure. PPE also places the entire burden of safety on the worker rather than on the system. That makes it the least reliable tier. It’s appropriate as a backup layer or as a temporary measure while better controls are being installed, but it should never be the primary line of defense when higher-level options are feasible.
How to Choose the Right Controls
OSHA recommends a structured process called a Job Hazard Analysis to identify which controls make sense for a given task. The steps are practical and designed to be done collaboratively with the people who actually do the work:
- Involve employees. The people performing a task usually know its dangers better than anyone observing from outside.
- Review accident history. Look at past injuries, illnesses, near misses, and equipment damage to identify patterns.
- Conduct a preliminary job review. Talk through known hazards with workers and brainstorm control ideas together.
- Rank and prioritize. List hazardous jobs and rank them by how likely they are to cause harm and how severe the consequences would be.
- Break the job into steps. Watch someone perform the task and write down every step, then identify where hazards appear within that sequence.
Once you’ve mapped the hazards, you work through the hierarchy from the top down. Can the hazard be eliminated? If not, can it be substituted? If not, what engineering controls could reduce exposure? Layer administrative controls and PPE on top as needed.
Why Controls Fail
Even well-designed hazard controls break down over time. The most common reasons come down to human factors. Organizations sometimes treat workers as though they’ll always be alert, always follow procedures, and always intervene correctly in an emergency. That’s not realistic. People make unintentional errors (slips and lapses), and they sometimes deliberately deviate from procedures (violations). A risk assessment that ignores these realities is incomplete.
Other common failures include choosing controls that are too low on the hierarchy when better options exist, failing to maintain engineering controls so they degrade over time, and never checking whether a control is actually working after it’s been put in place. Effective hazard control isn’t a one-time project. It requires periodic review: tracking injury rates, inspecting equipment, observing work practices, and asking workers whether the controls in place are actually practical to use on a daily basis.
Combining Multiple Layers
The strongest hazard control programs don’t rely on a single tier. They stack controls so that if one layer fails, another is still in place. A construction site might eliminate work at heights where possible, install guardrails where height work is unavoidable (engineering), train workers on fall prevention (administrative), and require harnesses as a final backup (PPE). Each layer catches what the one above it misses.
This layered approach is especially important for high-consequence hazards, where a single failure could result in a fatality or permanent disability. The more severe the potential outcome, the higher up the hierarchy your primary control should sit. Relying on PPE alone for a life-threatening hazard is a sign the control plan needs rethinking.