A score is only as trustworthy as the process behind it. This page is deliberately technical, because the difference between a review you can rely on and a review you cannot usually comes down to whether the reviewer tells you exactly what they measured, with what, and where the method stops being reliable.
ChairGrade evaluates every chair with the same four-stage protocol: a sixteen-point physical audit, a minimum hundred-day live-use period, a weighted six-criterion scoring pass, and a clinical review by a licensed physiotherapist. Nothing is published until all four stages are complete. What follows is the detail — tools, procedures, and the boundaries of our certainty.
The sixteen-point physical audit
Before a chair is sat in for a living, it is measured. Our audit records sixteen dimensions and characteristics, and the same order is followed every time so that two chairs can be compared without interpretation. The audit covers:
- Seat height range under load. Measured with a standardised 75 kg load applied to the seat, from the lowest to the highest usable position. Manufacturers quote figure-travel at rest; sag under load is what your body experiences.
- Seat pan depth. Front edge to backrest, at the seat's midpoint. This governs whether the chair supports your thighs without pressing behind the knee.
- Seat pan width. The usable sitting width at its narrowest point, not the widest outer dimension.
- Seat cushion thickness and compression. Measured at rest and under load, to calculate how far the foam yields and whether it bottoms out.
- Backrest height and shape. Total height and the vertical profile, including whether the upper back and shoulders are supported or left free.
- Lumbar crest position. The height at which the lumbar support's apex sits above the seat, both at its lowest and highest adjustment.
- Lumbar support depth. How far forward the lumbar region can be driven relative to the backrest plane, measured in millimetres.
- Lumbar firmness. A qualitative but consistent assessment of whether support is rigid, progressive, or soft, since the same geometry at different firmness behaves quite differently.
- Armrest height range and travel. Vertical adjustment plus any lateral, forward-back, and rotational movement, expressed as usable ranges rather than nominal ones.
- Armrest height relative to the seat. The critical figure — armrests that sit too high force the shoulders up; too low and the elbows cannot rest.
- Recline range. The angle of backrest travel, with and without any locking positions, and the amount of tension adjustment available.
- Headrest presence, reach, and depth. Recorded only where fitted, with the range of movement it offers.
- Base and caster specification. Base diameter and material, and caster type — hard-floor or carpet, braked or free, and whether they are interchangeable.
- Frame construction and materials. Where visible or disclosable, the materials used in the spine, backrest carrier, and seat plate.
- Weight and rated capacity. The chair's own mass and any stated maximum user weight, checked against the gas-lift class where identifiable.
- Gas lift classification. The cylinder class, since this is the single component whose failure is both most common and most consequential.
Where a dimension cannot be measured to a tolerance of ±5 mm, we say so and do not present it as precise. Where a manufacturer's published figure disagrees with our measurement, we record both and note the discrepancy, because that gap is itself useful information.
The instruments and the tolerances
We use a deliberately modest kit, and we name it so that our method can be examined rather than merely believed. Distances are taken with a steel rule and a digital caliper; heights under load with a marked vertical scale and a calibrated reference mass; angles with a digital inclinometer; and firmness by consistent hand-compression against reference samples. We do not own a materials-testing laboratory, and we do not pretend that our durability work is certification testing. It is instrumented observation of the kind a careful engineer would perform, documented so that its limits are visible.
Two tolerances matter most. For dimensional measurements we treat ±5 mm as the honest envelope of our tools. For anything described as a "range" — seat height, armrest travel, recline — we measure the usable span, not the nominal one, and we state it as the figure a user can actually reach. A chair with a 60 mm gas lift quoted as "120 mm of adjustment" because the figure is counted twice is a chair we will describe accurately.
The hundred-day live-use protocol
A measurement tells you what a chair is; only time tells you what a chair does. Every model in our evaluation is placed into active daily use for a minimum of one hundred days before a rating is published. This is not a staged sitting session with a stopwatch. It is a real desk, a real working day, and a real person who forgets the chair is a subject and simply uses it.
The length is not arbitrary. Most seating faults emerge slowly. Mesh loses tension over hundreds of load cycles and can begin to sag by month three. Molded foam compresses and either recovers or fatigues, and fatigue is what produces the hard, downhill seat that is comfortable in a showroom and punishing by mid-afternoon. Recline mechanisms develop play, casts shed their grease, and armrest pads deform where the elbows rest. A hundred days is long enough for these behaviours to surface and be documented, and short enough that our evaluations remain current.
During this period we record a structured daily log: comfort in the first hour, comfort after four hours, any pressure point reported by the sitter, noise from the mechanism, and any change in the chair's adjustment behaviour. Those notes, not just the final score, are what allow us to explain why a chair fails or endures in the written review.
Durability and wear assessment
Because we are not a certification laboratory, our durability work leans on two complementary approaches. The first is objective: we verify the components we can identify — gas-lift class, base material, caster construction, mesh weave — against what the manufacturer claims and what the market expects at that price. The second is experiential: the hundred-day log, plus a longer-term retention record for chairs that remain in our hands, documents how each model ages in the areas where failures actually occur.
Where a chair has survived attrition over months of use without measurable degradation of its seat surface, lumbar firmness, or mechanism smoothness, we say so and treat it as evidence of durability. Where it has not, we say that too. We are explicit that this is not the same as passing a multi-thousand-cycle certification test, and we do not present it as such. It is the honest, repeatable evidence a small independent team can gather, and for a reader deciding what to buy, it addresses the question that matters: will this still be good in three years?
The six weighted criteria
Scoring happens last, and against weights fixed before testing so that a striking chair cannot bend the rubric in its favour. Each criterion is scored out of ten, weighted, and combined into a single total. The weights are:
- Back and lumbar support — 25%. The single heaviest factor, reflecting the criteria's purpose. Assessed on lordotic curve preservation, lumbar height adjustability, sacral support, and the behaviour of the backrest through recline.
- Seat comfort and foam — 20%. Pressure distribution across thighs and sit bones, foam density and recovery, breathability, and the all-important question of whether it bottoms out under sustained load.
- Adjustability and range — 20%. Seat depth, armrest articulation, recline tension, tilt behaviour, and how well the chair adapts across a range of body heights rather than one.
- Build and materials — 15%. Frame rigidity, component quality, gas-lift class, base and caster construction, and the fit and finish that predict how the chair will age.
- Long-term durability — 10%. Evidence from the retention record and the hundred-day log about how the chair wears.
- Value within class — 10%. Price against performance, judged only within a comparable band, so that a budget winner is not punished for not being a premium chair.
Partial marks are awarded on a defined scale, and ties are broken in a fixed order — first by the lumbar-support sub-score, then by adjustability, then by durability — so that the resolution of a near-tie is not a matter of taste. The full scoring rubric, including the sub-criteria behind each weight, is published on our review methodology page.
Clinical review
The final stage before publication is clinical. A licensed physiotherapist examines the chair's postural behaviour against its measured geometry: whether the lumbar support actually lands at the height a spine requires, whether seat depth permits a neutral knee angle for the stated height range, and whether the recline encourages supported relaxation rather than a slumped collapse. This review can and does change scores, and it occasionally holds a review back entirely. Its scope and its limits are described on the medical review page.
Standardising the test environment
Measurements taken in a variable environment are not measurements. So the live-use stage is conducted against a fixed configuration wherever our testers work: a desk surface held at a consistent sitting height, a single monitor at a standardised viewing distance, a firm floor for the caster assessment, and a recorded room temperature, because both foam and mesh behave differently when a room is cold or humid. Where a chair has been tested in more than one location, we note it and check that our dimensional readings agree before merging the data.
This standardisation matters most for the subjective half of the assessment. Two testers who sit at different desk heights, or with differently angled monitors, are effectively testing two different sitting postures. By holding the workstation constant, we are able to attribute a reported pressure point or shoulder strain to the chair rather than to an incidental piece of furniture around it. It is a small discipline, but it is the difference between a controlled observation and an anecdote.
Comparing chairs across body sizes
An ergonomic chair is not ergonomic in the abstract; it is ergonomic for a body. So while our headline score is a general summary, our measurements are also tabulated against body-size bands so that a reader can see where a chair fits well and where it does not. We use the recognised approach of designing for the broad middle of body dimensions — roughly the fifth to ninety-fifth percentile in the relevant anthropometric measures — and we flag, explicitly, the chairs that sit at the edges of that band.
In practice this produces findings the headline score alone would hide. A chair with an unusually deep seat can score strongly overall yet be a poor match for a shorter sitter, because their knees cannot reach a neutral angle. A chair with a very high maximum seat height can suit a tall user beautifully and leave an average-height user with their feet off the floor. A chair with a tightly tuned lumbar apex is excellent for one torso length and intrusive for another. We publish these observations as fit notes, because buying decisions live in the detail, not in the average.
How to read a ChairGrade score
A ChairGrade score is a summary, and summaries are useful only if you know what they compress. The single number tells you how a chair performed against our weighted criteria for a general sitter at a given price. It does not tell you whether the chair fits you, and it is not designed to. For that, read the underlying measurements first and then the sub-scores, because a chair's sub-scores often reveal more than its total.
Two chairs can share an 8.2 and be entirely different products: one may earn it through exceptional lumbar support and modest adjustability, the other through broad competence and nothing outstanding. If your priority is lumbar geometry, the first is your chair; if you need the widest adjustment range for a team of different body types, the second may serve you better despite the identical number. We publish the sub-scores and the raw measurements precisely so that you can make that judgement yourself, and we would rather you did. A reader who understands why a score is what it is has no need to take our word for anything.
What this method does not do
Honesty about limits is part of the method. Our protocol does not certify a chair against any standard; we are not a test house, and no ChairGrade rating should be read as a compliance mark. It does not replace a fit assessment against your own body — the most important variables are your height, torso length, and desk, and no general score can substitute for checking measurements against your own. It does not evaluate every version of a product sold in every region; where specifications differ by market, we name the variant we tested. And it cannot predict an individual's comfort with certainty, which is exactly why we publish the underlying measurements alongside the score, so that you can second-guess us intelligently rather than take our word.
If this sounds like more process than a chair review deserves, we would argue the opposite. Eight thousand hours is a long exposure to a piece of furniture chosen on adjectives. The process is the point.