How to choose an office chair

Almost every “best office chair” page ends in a score. A number out of ten, a row of stars, a “9.2 — Editor’s Choice” badge. The trouble is that the score is for a body, and the body is not yours. Comfort happens where a specific chair meets a specific spine, and no rating on a page you are reading in another city can know how far your knees sit from the front of the seat, or where the curve of your back actually falls.

So this page does not score chairs. It hands you the checklist a score is standing in for — the handful of numbers that actually decide whether a chair fits you — in the order that matters, and it tells you where each one comes from. Get these right and the chair fits. Get them wrong and no amount of foam padding rescues it.

Most of the checklist is geometry, and geometry is honest: it either matches your dimensions or it does not. We lean on the one source that describes that geometry without trying to sell you anything — OSHA’s Computer Workstations eTool — and on the fact that the specs worth ranking are ranges, because a range is the only thing that tells you a chair can be moved to fit you, rather than hoping it happens to arrive at your size.

Why the range, not the rating

Here is the reasoning the rest of the page rests on. We have not sat in these chairs. We cannot tell you how the foam feels on the eighth hour, or whether a mesh edge bites the back of your legs, because that is a sensation and it belongs to you, not to us. Anybody who publishes a comfort verdict for a chair they shipped to a warehouse and photographed is inventing the one thing a reader most wants to be true.

What we canread off a spec sheet is whether the chair moves far enough to reach your body. A seat that adjusts from 16″ to 21″ can be set to a wider range of legs than one fixed at 18″. A seat pan that slides can clear the back of more knees than one that does not. The adjustment rangeis the honest, checkable version of “will it fit me” — it is published, and it is the thing we rank on in the best ergonomic office chairs. The rest of this page is the checklist of ranges to check, in order.

1. Seat height, and the fit test

Start at the floor, because everything above it is set by where your feet land.

OSHA describes a simple fit testfor seat height: raise or lower the seat until your feet rest flat on the floor and your thighs sit roughly parallel to it, which for most people lands the knees at something close to a right angle. That is the whole target — feet supported, thighs level, no pressure ridge under the thighs from a seat set too high, no knees-above-hips huddle from one set too low. It is the neutral seated posture the same eTool sketches out: feet down, thighs parallel, head balanced over the spine.

Which means the spec to check is the seat-height range, not a single figure. A typical gas cylinder gives you a few inches of travel; the question is whether your fit-test height falls inside it. Short and tall bodies are the ones this catches. If the lowest the seat goes still leaves your feet dangling, the chair is too tall for you and a footrest is the patch; if the highest it goes still buries your thighs, it is too short and there is no patch. We work through the actual movements on how to adjust your chair — but you cannot adjust past the end of the range, which is why the range is the thing you are really buying.

One aside that saves grief later: seat height is set to fit you, and then the desk is brought to the chair, not the other way round. If your fit-test height leaves the desk too high, that is a desk-and-keyboard-tray problem to solve above the seat — not a reason to perch the chair too high and let your feet leave the floor.

2. Seat depth, the number nobody checks

Seat depth is the distance from the backrest to the front edge of the seat, and it is the spec most likely to be wrong and least likely to be checked. It decides one thing: whether there is clearance behind your knees.

OSHA’s guidance is that when you sit back against the backrest, the seat should leave a gap — roughly a couple of fingers — behind the knees, so the front edge does not press into the soft tissue there. Too deep and you either perch forward off the lumbar support to save your knees, or you sit back and let the edge dig in. Too shallow and your thighs overhang the front unsupported.

This is why a sliding seat pan — the spec sheets call it “seat depth adjustment” — matters more than its plain name suggests, and why it is worth more to a very tall or very short person than almost any other feature. A fixed seat is built for one leg length. If yours is not that length, the chair fights you all day and no lumbar setting fixes it, because you are sitting in the wrong place relative to the back. Long-legged and short-legged buyers should treat seat depth as a first-tier spec, not a footnote — it is one of the specs we weight heavily in the best chairs for sitting all day.

3. Lumbar support that meets your actual curve

“Lumbar support” is the most abused phrase in the category, because a foam bump moulded into a backrest counts as lumbar support on a spec sheet whether or not it lands anywhere near your lumbar.

The lumbar is the inward curve of your lower back. Support that helps is support that meets that curve where it actually sits on yourspine — which means the spec that matters is not “has lumbar support” but “is the lumbar adjustable, in height and preferably in depth.” A pad you can raise, lower, and push in or out can be moved to your curve. A moulded bump is set to the designer’s idea of an average back, and if you are not that average it is pressing on the wrong vertebra all day.

We pull this apart in what lumbar support actually does, but the buying rule is short: treat “adjustable lumbar” as the real feature and “contoured backrest” as decoration until proven otherwise. As with everything on this page, it is the range of movement that tells you the chair can find your back, rather than hoping it was built for it.

4. Armrests that carry the forearm without the shrug

Armrests have exactly one job, and it is easy to test the wrong way. They are there to take some of the weight of your forearms so your shoulders do not have to hold your arms up all day. OSHA’s framing is that they should support the forearms without forcing you to shrug — if the armrest is too high your shoulders ride up to meet it, and if it is too low you lean to one side to use it, and both undo the point.

So the spec is, again, a range: armrests that adjust in height at a minimum, and ideally in and out (width) and forward and back (depth) too — the “4D” label. There is a second, sharper test that OSHA’s keyboard guidance hands you, because the armrest and the keyboard want to agree: “Elbows should be about the same height as the keyboard and hang comfortably to the side of the body.” Set the armrest so it carries the forearm at the same height your hands meet the keys — no shrug, no reaching down. Fixed armrests, or ones that only travel up and down, are the most likely to fail this across a range of body sizes. Armrests that are too tall and will not lower are a genuine reason not to buy a chair, because they force the exact shrug you were trying to avoid.

5. Weight capacity, the number to read first

This is the least glamorous spec and the one to check before any of the others, because it is a hard limit rather than a preference. A chair’s stated weight capacity is the load the manufacturer designed and tested the gas cylinder, base, and seat to carry. Sit above it and you are not “probably fine” — you are outside the envelope the chair was rated in, and the cylinder and base are precisely the parts that give way when it is exceeded.

Read the number, and read what it is attached to. A “big and tall” chair is not just a marketing tier; it usually means a higher-rated cylinder, a wider and deeper seat, and a reinforced base, and those are the parts that actually carry the rating. If a listing does not publish a capacity at all, treat the silence the way we treat every missing number on this site — as a gap, not a reassurance. This is one of the three specs we insist a chair publish before it can rank, and it is the one with the least room for interpretation.

6. Warranty, the longevity number

The last spec on the checklist is the one that stands in for “how long will this still be a chair.” Every product on this site carries three numbers, and the third is always the longevity number — for a chair, that is the warranty.

A warranty is not a comfort promise and it is not a quality score, but it is the closest thing to a manufacturer betting its own money on how long the mechanism lasts. Read the length, and read the structure: a headline “12-year” figure often covers the frame, while the parts that actually wear — the cylinder, the tilt mechanism, the arms, the upholstery — carry shorter terms buried underneath it. The moving parts are the ones that fail, so the warranty on the moving parts is the one that tells you something. How that longevity maps to what you should spend is the whole of how much to spend on a chair.

What BIFMA X5.1 actually is (and is not)

Somewhere in the shopping you will meet the letters BIFMA, usually deployed as if they were a grade. They are not. ANSI/BIFMA X5.1 is a test standardfor general-purpose office chairs — a published set of load, cycle, and stability tests a chair is put through: weights dropped and held on the seat, the backrest cycled tens of thousands of times, the base and legs loaded, the whole thing checked for tipping. A chair either passes those tests or it does not.

What passing means is narrow, and worth stating plainly: the chair survived a standardised durability and stability protocol. What it does not mean is that the chair is comfortable, that it fits you, or that it is “better” than a chair that was never submitted — testing costs money, plenty of good chairs skip it, and meeting X5.1 tells you nothing about whether the lumbar lands on your curve. So read “BIFMA certified” as “tested to a durability standard,” which is genuinely reassuring about the frame and cylinder, and not as a quality star. BIFMA’s own overview of its standards, and the full text of X5.1, are both public if you want to see exactly what the tests are; we have linked them below.

The order to check them in

Put together, the checklist runs in a deliberate order:

1. Weight capacityfirst — it is a hard limit, and if you are over it, nothing else matters.
2. Seat-height range— does the fit test (feet flat, thighs level, roughly a right angle at the knee) fall inside the travel?
3. Seat depth— is there clearance behind your knees, and does the pan slide if your legs are long or short?
4. Lumbar adjustability— can the support move to your curve, or is it a fixed bump?
5. Armrest adjustability— can they carry your forearms at keyboard height without a shrug?
6. Warranty— how long, and on which parts?

Notice what is not on the list: comfort, “ergonomic” as a bare adjective, and the brand’s own score. Those are either sensations we cannot measure for you, or claims a spec sheet cannot back. If you are still choosing between two named categories rather than two chairs, task chair vs office chair is the split to read first; once you have a shortlist, the best ergonomic office chairs ranks them on exactly these ranges.

What this page is not

This is not medical advice, and it is not written by anyone qualified to give it. Stephen is a home-office enthusiast who reads standards documents and spec sheets — not a physiotherapist, an ergonomist, or a clinician. Everything above is geometry: where a seat sits, how far a backrest moves, what a published test covers. It describes how a chair fits a body, not how to treat one.

We have not sat in these chairs, and we have not measured your spine. If sitting causes you pain that persists, the person to see holds a licence and can examine the back that an affiliate page never can — no adjustment range on any spec sheet is a substitute for that. What this checklist can do is stop you buying a chair that cannot physically be set to your dimensions, which is a smaller promise than the star ratings make, and a truer one.