Hazard control is any action taken to eliminate, reduce, or limit exposure to something that could cause injury or illness in a workplace. It ranges from removing a dangerous chemical entirely to handing a worker a pair of safety goggles. The guiding principle is simple: the less a control depends on human behavior, the more reliable it is. That idea shapes every serious safety program in every industry.
The Hierarchy of Controls
Safety professionals don’t treat all controls as equal. NIOSH, the federal agency that researches workplace safety, ranks controls in a five-level hierarchy based on how effectively they protect people. From most effective to least effective, the levels are: elimination, substitution, engineering controls, administrative controls, and personal protective equipment (PPE).
The hierarchy is meant to be followed from top to bottom. You start by asking whether you can eliminate the hazard entirely. If not, you move down to substitution, then to engineering solutions, and so on. In practice, most workplaces use a combination of controls at different levels, but the goal is always to push protection as high up the hierarchy as possible.
The top three levels (elimination, substitution, and engineering controls) are more effective because they work without requiring people to do anything differently. The bottom two (administrative controls and PPE) demand significant, ongoing effort from workers and supervisors, which means they’re more likely to fail on a busy Tuesday afternoon when someone is tired or distracted.
Elimination and Substitution
Elimination removes the hazard at its source. This could mean redesigning a work process so that a toxic chemical, heavy object, or sharp tool is no longer part of the job at all. No exposure can occur when the hazard simply doesn’t exist. A warehouse that replaces a manual lifting task with a conveyor system has eliminated a musculoskeletal hazard. A contractor who pre-fabricates components on the ground instead of assembling them at height has eliminated a fall hazard.
Substitution keeps the process but swaps the dangerous element for a safer alternative. Switching from a solvent-based cleaner to a water-based one, or replacing a silica-containing abrasive with one that doesn’t generate harmful dust, are common examples. Substitution is most practical during the design or planning phase of a project, when changes are cheap. Retrofitting an existing process is harder but still worth evaluating.
Engineering Controls
When you can’t get rid of the hazard, the next best option is to physically separate it from people. Engineering controls do this through structural or mechanical changes to the work environment. They reduce exposure by preventing hazards from reaching workers while still allowing those workers to do their jobs.
Common examples include:
- Local exhaust ventilation that captures fumes or dust at the point where they’re generated, before they reach a worker’s breathing zone
- Machine guards and interlocks that prevent a machine from operating when a worker’s hand is near moving parts
- Guardrail systems along elevated platforms or open edges
- Noise enclosures around loud equipment
- Lift equipment that removes the need to manually carry heavy loads
- Pedestrian barriers and designated aisles that separate foot traffic from forklifts and other vehicles
The key advantage of engineering controls is that they’re always working. A guardrail doesn’t need to remember to show up, and a ventilation hood doesn’t get complacent after six months on the job. Once installed and properly maintained, these controls provide continuous protection regardless of how many people rotate through the workspace.
Administrative Controls
Administrative controls don’t remove or block the hazard. Instead, they change the way people work so their exposure is reduced in duration, frequency, or intensity. Think job rotation schedules that limit how long any one person works near a noise source, or standard operating procedures that require a specific sequence of steps before entering a confined space.
Training programs, warning signs, restricted-access zones, and buddy systems all fall into this category. So do work-rest schedules in hot environments and shift limits for jobs involving repetitive motion. These controls are useful and sometimes necessary, but they rely on people consistently following rules. That makes them inherently less reliable than the levels above them. A sign that says “Wear Hearing Protection Beyond This Point” only works if every worker reads it, has protection available, and puts it on every single time.
Personal Protective Equipment
PPE sits at the bottom of the hierarchy. Hard hats, safety glasses, respirators, gloves, high-visibility vests, and fall harnesses are all forms of PPE. They don’t reduce the hazard itself; they put a barrier on the worker’s body. If every other layer of protection fails or isn’t feasible, PPE is the last line of defense.
It’s also the most failure-prone. PPE only works when it’s worn correctly, fits properly, and is maintained. Improperly sized equipment can be outright dangerous: oversized gloves can get caught in machinery, and a loose harness may not arrest a fall. A 2024 OSHA rule specifically addressed this problem in the construction industry, requiring employers to provide PPE that properly fits each worker who needs it, aligning construction standards with rules that already applied to general industry.
None of this means PPE is unimportant. In many situations it’s essential. But a workplace that relies primarily on PPE without seriously evaluating higher-level controls is accepting more risk than it needs to.
How Hazard Controls Are Selected
Choosing the right control starts with hazard identification and risk assessment. You first figure out what hazards exist, who is exposed, how often, and how severe the consequences could be. That assessment then directly informs which controls are appropriate and where to invest resources.
A high-severity, high-frequency hazard warrants elimination or engineering controls. A lower-risk hazard that workers encounter rarely might be adequately managed with administrative procedures and PPE. Cost is a legitimate factor, but safety regulations generally don’t accept “too expensive” as a reason to skip higher-level controls when they’re feasible. The expectation is that you work down from the top of the hierarchy and document why you landed where you did.
Most real-world safety programs layer multiple controls. A manufacturing facility might install machine guards (engineering), train workers on lockout/tagout procedures (administrative), and require cut-resistant gloves (PPE) for the same piece of equipment. Each layer compensates for the possibility that another layer fails.
Formal Safety Management Systems
For organizations that want a structured framework, the ANSI/ASSP Z10.0 standard provides one of the most comprehensive systems-based approaches to occupational safety management. It gives organizations an architecture for continuously identifying and eliminating risks, complying with regulations, and engaging both management and frontline workers in safety improvement. It’s designed to be customizable, so a 50-person fabrication shop and a multinational chemical company can both use it. A quick-start guide (the GM-Z10.101) is available as a free download for small and medium-sized organizations that want to start without committing to the full standard.
Whether or not you adopt a formal standard, the underlying logic of hazard control stays the same: identify what can hurt people, control it as close to the source as possible, and verify that your controls are actually working over time.