Manual Handling Risk Assessment: 9 Steps for EHS

HSE reports that handling, lifting, or carrying caused 17% of employer-reported non-fatal injuries to employees in Great Britain in 2024/25. This guide shows how an EHS manager can build a manual handling risk assessment that changes the task, not only the training record.

Why manual handling risk assessment fails in real operations

Manual handling risk assessment fails when it treats injury as a worker technique problem instead of a work-design problem. The hazard sits in load weight, reach distance, grip quality, repetition, posture, floor condition, time pressure, staffing, and supervision, which means a form that ends with refresher training has usually stopped too early.

HSE defines manual handling as transporting or supporting a load by hand or bodily force, including lifting, putting down, pushing, pulling, carrying, or moving. That definition matters because many companies assess only lifting, while the injury pattern is often created by pushing a stuck trolley, twisting with a box, dragging a hose, or carrying materials through a route that was never designed for people.

As Andreza Araújo argues in Safety Culture: From Theory to Practice, culture is visible in repeated decisions under pressure. In manual handling, that pressure appears when production accepts a heavy lift because the pallet is late, the hoist is unavailable, or the supervisor believes the experienced worker can manage it.

Step 1: Define the task exactly

A useful manual handling risk assessment starts with one defined task, one location, and one normal work cycle. The assessor should write down the load, route, frequency, duration, number of workers, equipment used, and the point in the shift when the task occurs, because risk changes when a 12 kg box is lifted 4 times a day versus 400 times in 8 hours.

Across 25+ years leading EHS at multinationals, Andreza Araújo has observed that vague task names hide the real exposure. A line called "materials movement" may include unloading, sorting, twisting, reaching above shoulder height, pushing a cart, and stacking at floor level, each with a different control logic.

The practical rule is to assess the job the worker actually performs, not the task as imagined by the procedure. Observe at least 3 cycles, including one busy period, then write the assessment around the physical sequence: pick up, carry, turn, place, return, and repeat.

Step 2: Which jobs should you assess first?

The first jobs to assess are the ones with injury history, discomfort reports, high repetition, awkward posture, heavy or unstable loads, long carries, and tasks performed by new or temporary workers. A site with 50 manual tasks cannot assess everything deeply in 1 week, so the first screen should rank exposure before paperwork volume.

EU-OSHA explains that work-related musculoskeletal disorders are linked to physical, organizational, psychosocial, and individual factors. That is why the first priority should not be only the heaviest item. A lighter load handled under pace pressure for 6 hours can be more harmful than a heavier load lifted occasionally with mechanical help.

Use 4 filters to set priorities: injury and first-aid history, worker discomfort, visible strain during observation, and production dependency. The fourth filter matters because a task done every shift by 30 operators deserves attention before a rare task that looks dramatic but creates little routine exposure.

4 filters are enough for a first triage, provided the EHS manager validates them with supervisors and workers before locking the assessment plan.

Step 3: Map the load, person, task, and environment

The assessment should map 4 dimensions: the load, the person, the task, and the environment. Those dimensions keep the assessor from reducing manual handling to a single weight limit, because injury risk also depends on reach, grip, asymmetry, floor condition, thermal stress, visibility, and fatigue.

NIOSH identifies force, posture, repetition, duration, contact stress, vibration, and temperature as risk factors that can increase work-related musculoskeletal disorder risk. For an EHS manager, that list is a reminder that the risk assessment must capture body position and exposure time, not only load mass.

Write each dimension in plain operational language. A weak entry says "awkward posture." A useful entry says "operator reaches 55 cm across the pallet, rotates trunk left, then places 16 kg cartons on a shelf at shoulder height for the last 2 hours of the shift." Specific descriptions make controls easier to design and easier to verify.

Step 4: Remove avoidable handling before scoring risk

The strongest assessment question is whether the manual handling can be eliminated before anyone scores it. ISO 45001:2018 requires organizations to identify hazards, assess OH&S risks, and apply controls, and the hierarchy of controls makes avoidance and engineering stronger than administrative reminders.

ISO specifies the requirements for an occupational health and safety management system, including risk assessment, worker participation, and continual improvement. In practice, this means the manual handling assessment should ask whether the load can be delivered closer, lifted mechanically, split, slid, palletized differently, or removed from the sequence.

This is where many programs become compliance theater. They score the task, prescribe training, and close the action while the same pallet height, same distance, and same production pressure continue. The stronger route is to connect manual handling to the hierarchy of controls and force a redesign question before the training answer appears.

Step 5: How do you involve workers without turning it into a complaint session?

Worker involvement works when the assessor asks for task evidence, not general dissatisfaction. The best questions are concrete: where does the load feel unstable, which part of the route makes you twist, when does the cart become hard to push, and what changes after the second hour?

In more than 250 cultural-transformation projects supported by Andreza Araújo's team, a recurring pattern appears: workers often know the exact point where the task becomes unsafe, although they stop reporting it when every answer becomes "be careful" or "use proper technique." That silence is cultural data, not attitude failure.

Use a 20-minute field conversation with 2 workers and 1 supervisor. Ask them to mark the worst moment of the cycle, then compare their answer with the observed posture, pace, and route. This keeps the dialogue practical and prevents the assessment from becoming either a blame exercise or an unstructured wish list.

Step 6: Rate risk only after controls are visible

Risk rating should happen after existing controls are checked in the field. A manual handling task may look acceptable on paper because a hoist, trolley, or team-lift instruction exists, but the actual risk remains high if the hoist is 40 meters away, the trolley wheels are damaged, or the second person is unavailable during peak demand.

The risk matrix is useful only when the assumptions behind the score are true. If the assessment states that mechanical aid is available, the assessor should verify condition, access, training, and use during normal production, because an unused control does not reduce exposure.

Connect the score to visible evidence. The same logic applies in a Take 5 safety check: the value is not the paper pause, but the worker's ability to see whether the control exists before the job begins.

40 meters between the task and the hoist can turn an available control into a fictional control, especially when production pace rewards the fastest workaround.

Step 7: Choose controls that change the physical exposure

The best controls change the physical exposure by reducing force, reach, repetition, asymmetry, distance, or duration. Training has a place, but it should support a redesigned task instead of compensating for a poor layout.

Control selection should follow a clear order: eliminate the lift, reduce load weight, improve grip, change shelf height, shorten the carry, add mechanical assistance, redesign packaging, adjust staffing, and then train the worker on the changed method. This order prevents the familiar mistake of asking the body to absorb a design problem.

Manual handling also interacts with traffic flow. If a pallet route crosses forklifts, uneven floors, or pedestrian congestion, the assessment should connect with the workplace traffic plan because pushing, pulling, stopping, and sudden turning can create both musculoskeletal and struck-by exposure.

Step 8: Assign owners, deadlines, and verification evidence

Every accepted control should have an owner, a deadline, and evidence that proves the exposure changed. Without those 3 elements, the assessment becomes a record of concern rather than a management tool.

As Andreza Araújo writes in Safety Culture Diagnosis, diagnosis only creates value when it is converted into an execution plan that leaders can carry out and verify. Manual handling follows the same principle: the finding is not the improvement, and the recommendation is not the control.

Use a simple action standard. Engineering changes need a responsible manager, budget decision, target date, and commissioning check. Administrative changes need the revised method, communication record, supervisor observation, and worker input after 30 days. Training changes need evidence that the task itself is controlled, not only that attendance was recorded.

Step 9: What should the final checklist contain?

The final checklist should prove that the assessment covered task reality, exposure priority, worker input, control design, risk rating, ownership, and follow-up. A checklist is not a substitute for judgment, although it prevents common omissions when several supervisors assess similar jobs.

Use the checklist below before closing the assessment:

• The task name identifies one location, one activity, and one normal work cycle.
• The assessment separates load, person, task, and environment factors.
• At least 2 workers and 1 supervisor contributed field evidence.
• Elimination or mechanical assistance was considered before training.
• The risk score matches verified controls, not assumed controls.
• Each action has an owner, date, and verification method.
• The reassessment date is triggered by change, injury, discomfort report, or 90-day review.

Every month that a manual handling task remains unchanged after discomfort reports creates a preventable injury pipeline, while the organization keeps proving that it heard the signal and chose not to redesign the work.

Comparison: training-only vs control-led assessment

A control-led manual handling assessment changes the task before it lectures the worker. The difference is visible in what the EHS manager records, what leaders fund, and what supervisors verify after the form is signed.

The control-led path is more demanding, but it is also more honest. It recognizes that back injuries, shoulder strain, and chronic discomfort are not solved by a poster when the physical system still asks people to lift, twist, push, and repeat beyond what good work design should allow.

Conclusion: assess the work, not only the worker

Manual handling risk assessment becomes credible when it changes exposure through elimination, engineering, layout, pace, ownership, and verification.

If your organization wants to move from lifting reminders to real task redesign, Andreza Araújo's work in Safety Culture: From Theory to Practice, Safety Culture Diagnosis, and ACS Global Ventures consulting offers a practical route from diagnosis to operational change. Start the conversation with Andreza Araújo at Andreza Araújo.