Skip to content
Desk and Daylight

Desk height calculator

Every other calculator asks for your body height. No standard supports that. This one asks for your elbow height, because that is what the arithmetic actually needs — and it shows you the working.

By Stephen V.Last reviewed

Almost every desk-height calculator on the internet asks for your body height and hands back a number. We went looking for the standard that justifies that, and there isn’t one.

What exists is ANSI/HFES 100-2007, “Human Factors Engineering of Computer Workstations” — a genuine ANSI-accredited US national standard. We read it. Every equation in it anchors on elbow height. Its instruction for a standing workstation is, word for word, “Place the support surface at standing elbow height.” Body height does not appear in the chain at all.

So this calculator asks for your elbow height. It takes thirty seconds to measure and it is exact for your body rather than for a population average.

Work out your desk height

The standard this is built on anchors on your elbow height, not your body height. Measuring takes thirty seconds and is exact for you. The estimate is a fallback, and we’ll show you its working.

Which ANSUR percentile table to draw the leg-length figure from.

Sit upright, feet flat, upper arms hanging, elbows at 90°. Measure from the seat to the underside of your elbow.

Enter a measurement above and the number — and the arithmetic behind it — appears here.

What the standard actually says

HFES 100-2007 does not leave desk height to taste. It sets mandatory ranges, using the word shall:

  • Seated (§8.3.2.4.1): the input-device support surface shall have height adjustment covering at least 56–72 cm (22–28.3″), floor to top of surface.
  • Standing (§8.3.2.4.2): place the surface at standing elbow height, adjusting between 95–118 cm (37.4–46.5″).
  • Sit/stand (§8.3.2.4.3): adjust between 56–118 cm (22–46.5″), and cover some portion of the seated range.

Notice the top of the seated range: 28.3″. A conventional fixed desk is 29″. The standard’s entire seated range sits below the desk most people own.

The arithmetic, in full

HFES 100’s Equation 8-1 gives input device height as A + sin(B) × C, where A is standing elbow height, B is the elbow angle relative to horizontal, and C is elbow-wrist length. With the elbow at horizontal, the sine term is zero and the whole thing reduces to: put the surface at elbow height.

For seated work the chain runs through popliteal height — the distance from the floor to the back of your knee. ANSUR defines seated elbow rest height as measured from the sitting surface, so adding popliteal height puts the number back on the floor:

seated desk height = seated elbow rest height + popliteal height

That assumes your chair is at your popliteal height and your feet are flat, which is the reference posture the standard describes and the one OSHA describes too: “Knees are about the same height as the hips with the feet slightly forward.”

Where the numbers come from

The population figures are from ANSUR II, the 2012 Anthropometric Survey of U.S. Army Personnel — 93 measurements on 6,068 people, released publicly in 2017. We downloaded the raw CSVs and computed the percentiles rather than copying a table from someone else. Here are the medians, in centimetres:

  • Male(n=4,082): stature 175.5 · seated elbow rest 24.6 · popliteal 43.0
  • Female(n=1,986): stature 162.6 · seated elbow rest 23.3 · popliteal 38.7

Which gives, by the equation above: 67.6 cm (26.6″) for the median man and 62.0 cm (24.4″)for the median woman. Both below 29″. That is our arithmetic, not a published figure, and we are labelling it as such.

The honest limits of this

Three of them, and they are real.

ANSUR is a military population.Younger and fitter than a general home-office audience. It is the only public dataset that contains the two dimensions this calculation needs — we checked whether CDC’s NHANES could substitute, and it cannot: the words “popliteal”, “elbow rest” and “seated elbow” appear zero times in its anthropometry procedures manual. NHANES simply does not measure them. So ANSUR is what there is, and its population skew is a limitation we cannot engineer away.

The estimate path is a proportional model, not a standard. If you use the height-based option, we scale ANSUR’s medians by your stature. Proportions vary independently of height. It is better than nothing and worse than a tape measure, and the calculator says so on screen rather than in a footnote.

We read HFES 100-2007 from a third-party copy(a vendor’s hosted PDF, linked in our sources), not from HFES or ANSI directly, and we have not confirmed whether 2007 remains the current edition. The content is self-evidently the standard, but you should know where we got it.

What we deliberately did not use

ISO 9241-5is the correct ISO part for workstation geometry, and it is paywalled. We verified its title, scope and 2024 edition from ISO’s own catalogue and read nothing else. So no number on this page is attributed to it, and you should be suspicious of any site that quotes ISO dimensions without saying which edition they bought.

ANSI/BIFMA G1-2013, the ergonomics guideline, is also paywalled. The figures you will see attributed to it — “22-30 inches seated”, “95th percentile male = 6 ft 3 in” — trace to vendor blogs rather than to BIFMA. We could not verify any of them, so we have not repeated them. Worth knowing that G1 is a guideline, not a national standard: it has not been through the ANSI review process that HFES 100 has.

If your desk can't reach your number

This is the common case, especially for shorter people, and it has three honest answers.

Raise the chair and support your feet. Free, and it works. Set the chair so your elbows meet the desk you already have, then put your feet on something solid so they are not dangling. You have moved yourself to the desk instead of the desk to you, which is not ideal but is a great deal better than typing with your shoulders up.

Add a keyboard tray.The desk stays where it is and the input surface — which is the thing the standard actually specifies — drops to your height.

Get a desk that goes low enough. Note that this is the opposite of how standing desks are sold. Every listing advertises how high it goes; the floor height is what decides whether it fits you seated, and it is the spec almost nobody prints. Across the four desks in our standing desk roundup the bottom of the range runs from 21.6″ to 29.2″ — nearly eight inches of difference between desks that look identical in a photograph.

Common questions

How high should a desk be?

There is no single right answer, and any page that gives you one number is guessing. The honest version: your desk should sit at your seated elbow height, with your feet flat and your elbows at roughly 90°. For the median adult in the ANSUR II data that works out around 67.6 cm (26.6″) for men and 62.0 cm (24.4″) for women — both belowthe 29″ a standard desk gives you. Use the calculator above with your own measurement rather than either of those.

Does OSHA say what height a desk should be?

No — and this is worth knowing, because plenty of pages claim it does. OSHA’s Computer Workstations eTool gives posture principles and exactly one number, and that number is leg clearance: “Clearance for the legs, under the desktop, should generally be between 20-28 inches (50-72 cm) high.” That is the space for your knees, not the height of the surface. OSHA prescribes no desk height at all, and the eTool is advisory guidance rather than a regulation. Anyone telling you “OSHA says 29 inches” has invented it.

Why does this ask for my elbow height instead of my height?

Because that is what the standard uses. ANSI/HFES 100-2007 anchors every one of its surface-height equations on elbow height — its instruction for the standing case is literally “Place the support surface at standing elbow height.” No authoritative standard publishes a formula that maps body height to desk height. The stature→elbow step that other calculators perform silently is a population correlation, and it is where all the error enters: two people the same height can have elbow heights several centimetres apart.

Where does the 0.4739 × height formula come from?

Nowhere we could find. That coefficient (and its partner, 0.5538) appears on Omnicalculator’s desk-height calculator and is reproduced across the web. We checked their page: it cites no source, study or standardfor either number. We could not trace them to any dataset. We are not saying they are wrong — we are saying nobody can show you where they came from, which is a different problem and arguably a worse one.

Is my standard 29-inch desk too tall?

For the median adult, on this arithmetic: yes. Both ANSUR medians land below 29″ — 26.6″ for men and 24.4″ for women. That is not a scandal; it is why keyboard trays exist, and why height-adjustable desks are the category they are. If you are on a fixed desk, the standard fix is to raise your chair until your elbows meet the desk and then support your feet, rather than lowering your arms to reach it. See the full setup order.

How accurate is the estimate-from-height option?

Less accurate than a tape measure, and we would rather you did not use it. It scales ANSUR’s median elbow and leg measurements by your height, which assumes your proportions match the median — and proportions vary independently of height. It is there because some people will not measure, and a labelled estimate with its arithmetic shown beats an unlabelled one. But measuring takes thirty seconds and removes the guesswork entirely.

Sources

Every figure on this page comes from one of these. If a manufacturer doesn’t publish a number, we print “—” rather than estimate it.

  1. ANSI/HFES 100-2007, Human Factors Engineering of Computer Workstations — §8.3.2.4 surface height requirements, Eq 8-1, Appendix A Eq A-6/A-7/A-8, Table A-1 (full text) — read 2026-07-16
  2. HFES technical standards — the publisher's own standard listing — read 2026-07-16
  3. OSHA Computer Workstations eTool — Desks (the 20–28″ leg-clearance figure; note OSHA prescribes no desk height) — read 2026-07-16
  4. OSHA Computer Workstations eTool — Good Working Positions (elbows 90–120°, reference postures) — read 2026-07-16
  5. ANSUR II (2012 Anthropometric Survey of U.S. Army Personnel) — OPEN Design Lab landing page — read 2026-07-16
  6. ANSUR II raw public CSVs — the percentiles on this page were computed from these — read 2026-07-16
  7. CDC NHANES Anthropometry Procedures Manual — evidence that NHANES does NOT measure popliteal or seated elbow height — read 2026-07-16
  8. Knoll / Dr. Michael O'Neill, 'Office Ergonomic Standards: A Layperson's Guide' — on why HFES 100 is a national standard and BIFMA G1 is not — read 2026-07-16
  9. ISO 9241-5:2024 — title, scope and edition only. The full text is paywalled and we have not read it, so no dimension on this page is attributed to it. — read 2026-07-16
  10. Omnicalculator's desk height calculator — the widely-copied formula that cites no source — read 2026-07-16

Read next