OCRA vs NIOSH vs REBA: the complete guide

Every ergonomic assessment starts with a choice that is easy to get wrong: which method to use. Pick the wrong one and the score you get back is precise, defensible-looking and beside the point. A lifting task scored with a posture method will miss the load; a repetitive assembly task scored with a lifting method will miss the repetition that is actually wearing out the operator’s shoulder.

This guide is about that choice. It covers the three methods that industrial teams use most: the Revised NIOSH Lifting Equation for manual lifting, the OCRA method for repetitive work of the upper limbs, and REBA for whole-body postures. For each one you will find what it was designed for, the data it needs, how the score is built and how to read it — and then a way to choose between them, three worked examples and the mistakes that most often skew results.

It is written for HSE managers, ergonomists and industrial engineers who run assessments on real lines, not in a lab. You do not need to know the methods already; if you do, chapters 2, 6 and 7 are where most of the practical value is.

What you need

• A short video of each task, filmed from the right and the left side. Two or three complete cycles are enough.
• The task data that video cannot give you: weights of the loads, cycle time, shift length, breaks and how many hours a day the task is done.
• Someone who knows the task: the operator or the line leader. Five minutes of conversation will tell you about the hard cycles that the video missed.

The three methods answer different questions. That is the most important thing to know about them.

NIOSH asks: how heavy can this lift be? It was developed by the US National Institute for Occupational Safety and Health to prevent low-back pain from two-handed lifting and lowering. It gives you a Recommended Weight Limit for the specific conditions of the task and compares it with the actual load.

OCRA — Occupational Repetitive Actions — asks: how much repetitive effort can the arms and hands sustain over a shift? It was developed in Italy by Colombini and Occhipinti for tasks that repeat the same actions in cycles or recurring patterns, such as assembly, packing and machine feeding, and it is the method referenced in ISO 11228-3 and EN 1005-5. It weighs how often the hands act, how hard, in which postures and with how much recovery time.

REBA — Rapid Entire Body Assessment — asks: how risky is this posture? It was published by Hignett and McAtamney in 2000 to score whole-body postures quickly, including the trunk, neck and legs, with corrections for load, grip and activity. It is a snapshot method: you score the worst or most frequent posture, not the whole shift.

Choose the method from what the operator actually does: NIOSH if they lift or lower loads with both hands, OCRA if the arms and hands repeat the same actions in cycles or recurring patterns, REBA if the problem is the posture of the whole body, Snook & Ciriello if they push or pull a load on wheels. Many stations need more than one.

The decision tree

Ask the four questions in order, for one task at a time. Every “yes” adds a method.

If none applies, screen the task with REBA to confirm, and reassess if the task changes.

Two clarifications that change the answer

• Repetitive does not mean machine-paced. An operator preparing orders of different sizes, with no fixed cycle time, does repetitive work if the actions are always the same. OCRA applies.
• Occasional handling is not a NIOSH task. In Postura’s practice, a load moved once every few months is treated as an injury risk, not biomechanical overload. Record it in the risk assessment with a short note instead of computing an index.

The Revised NIOSH Lifting Equation computes a Recommended Weight Limit (RWL) for the specific lift and compares it with the actual load. The ratio between the two is the Lifting Index (LI).

RWL = LC × HM × VM × DM × AM × FM × CM

LI = load lifted ÷ RWL

The load constant

The load constant (LC) is the heaviest load that can be lifted under ideal conditions. The NIOSH value is 23 kg, set to protect about 99% of men and 75% of women. ISO 11228-1 and EN 1005-2 allow reference masses adapted to the working population, such as 25, 20 or 15 kg.

The six multipliers

Each multiplier is 1 under ideal conditions and falls towards 0 as the lift gets worse.

The horizontal multiplier is often the one that dominates; the distance multiplier rarely does, because its range is narrow. The frequency multiplier is the only one that captures work organisation.

OCRA scores repetitive upper-limb work over the shift, for each side, in two instruments: the OCRA Checklist for mapping and prioritising, and the OCRA Index for detailed analysis. Its High-Precision version, the HP Checklist, builds the Checklist score from measured times.

Checklist or Index

Both are validated and assess the same risk: choose by what the output is used for.

The five factors

The Checklist adds five factors into a Partial score:

• Frequency: technical actions per minute, dynamic and static.
• Posture: time spent by shoulder, elbow, wrist and hand outside neutral; the highest of the joints counts.
• Stereotypy: very short cycles or the same actions repeated for most of the cycle.
• Complementary factors: gloves, vibration, impacts, cold, precision, and a work rate set by the machine.
• Force: perceived effort on the Borg CR-10 scale, with its duration in the cycle.

The Partial score is then multiplied by a recovery multiplier, from the hours without adequate recovery in the shift, and by a duration multiplier, from the net minutes of repetitive work.

REBA scores one whole-body posture from 1 to 15: it combines the posture of the trunk, neck and legs (Group A) and of the arms and wrist (Group B), corrects them for load, grip and activity, and maps the result to five risk levels.

Group A: trunk, neck and legs

The three scores are combined in Table A, then the force/load score is added: 0 below about 5 kg, +1 for 5–10 kg, +2 above 10 kg, and +1 for shock or a rapid build-up of force. The result is Score A.

Group B: arms and wrist

Group B is assessed one side at a time. The scores are combined in Table B, then the coupling score is added: 0 good, +1 fair, +2 poor, +3 unacceptable. The result is Score B.

Three common tasks, each assessed with the method it needs. The numbers are illustrative: they show how the inputs become a result, not the risk of any real station.

1. Picking from a low pallet: NIOSH

An operator lifts 12 kg boxes from a pallet to a conveyor, twice a minute, in blocks of up to two hours, each followed by at least 36 minutes of light work (medium duration). The boxes have fair hand holds.

RWL = 23 × 0.625 × 0.835 × 0.895 × 0.904 × 0.84 × 0.95 = 7.75 kg

LI = 12 ÷ 7.75 = 1.55: light red, low risk. The horizontal and vertical multipliers drive it: raising the pallet and bringing the boxes closer would do more than any other change.

Most unreliable scores come from the input, not from the method. Five mistakes account for most of them, and every one of them is invisible in the result.

1. Estimating the weight

The load enters NIOSH directly as the numerator of the Lifting Index. A weight estimated by the operator, or copied from an old assessment, goes straight into the risk band, and no step of the calculation can detect it. Weigh the loads and count the lifts over a representative period.

2. Recording a segment that is not the cycle

A clip that covers two pieces declared as one cycle doubles the declared cycle time, and everything computed from it follows. A clip that stops before refills or packaging leaves out the lighter part of the work and scores only the dense part. Find the first and last frame of one complete repetition before you compile.

3. Filming only the worst condition

If a station runs light and heavy variants and the video shows only the heaviest, the analysis treats it as the whole job and overstates the risk. Decide which conditions represent the station before you record.

4. Cutting acyclic time out of the task

Restocking, machine time and waiting inside a repetitive task are part of it. Moving them to non-repetitive time leaves only the dense part and overestimates the OCRA score. Likewise, the cycle times of a time-and-methods study include allowances and line saturation: reconcile them with OCRA’s net repetitive time before comparing.

When a station combines several demands, assess each with its own method and report the results side by side: a NIOSH index for the lifts, an OCRA score for the repetitive arm work, a REBA level for the worst postures. Do not average them into one number.

One station, several tasks

Split the job into tasks: the parts of the shift with a stable pattern of work. A packing station might have a repetitive packing task, scored with OCRA, and a palletising task every hour, scored with NIOSH. Each task gets its own method, and each result says what it covers.

Rotations

If the operator rotates between several repetitive tasks, the score of each task is an intermediate result. The overall exposure comes from the OCRA multitask procedure for repetitive work and from the Sequential Lifting Index for lifting. Never report a single-task score as the exposure of someone who rotates.

Take one station you know well and run the decision tree in chapter 2. Write down which methods it needs and what data each one asks for. That list is the plan for your next assessment, and usually the first sign that one method was never going to be enough.

With Postura

From the videos of a station, Postura computes the OCRA High-Precision Checklist for repetitive work, with AI-counted technical actions and awkward postures, and REBA for whole-body postures where it is enabled on your account. NIOSH (lifting) and Snook & Ciriello (pushing and pulling) are computed from the weights, parameters and forces you measure and enter. Time and methods analysis runs on the same videos.

• See it on your own footage: book a demo.
• Going deeper on OCRA: read OCRA High-Precision Checklist.
• Pushing and pulling: read Pushing and pulling.
• Recording the video: read How to film a workstation for an ergonomic assessment.