Work and health

Ergonomic desk set up: what the evidence says, and what to change today

A practical guide for anyone at a computer most of the day, in an office or at home, with every claim graded against the research.

Highland Willow · Boone, North Carolina · 8 November 2026

Start here: fix these today, in this order

You can do all six in about 20 minutes with a tape measure and whatever books or boxes you have. The calculator further down this page gives starting heights for your body. They are a starting point to adjust to comfort, not exact figures, especially if you estimate from your height alone: height-based estimates use averages from the ANSUR II body-size survey. [Gordon 2014]

1. Set the chair first. Raise or lower it until your feet are flat and your thighs are about level, and the back of your knee clear of the seat edge: a gap of two to four fingers is a common rule of thumb. Sit all the way back so the backrest supports your lower back. [OSHA chairs; HFES 100 draft §5.19]
2. Bring the work up to your elbows, not your elbows up to the work. With your shoulders relaxed and your upper arms hanging, your elbows should be about level with the keyboard. [OSHA keyboards; CCOHS keyboard] Slightly above is fine: a more open elbow was linked to fewer neck and shoulder problems. [Marcus 2002] If the desk is higher than your elbows, raise the chair until they clear it, then put a footrest or a sturdy box under your feet.
3. Keep the keyboard flat. If its back feet tilt it up and bend your wrists back, fold them in. [OSHA keyboards] A keyboard tray that slopes slightly away from you kept wrists straighter in a small trial. [Hedge 1999]
4. Pull the mouse in. Put it right beside the keyboard, at the same height. Reaching out to the side for a mouse loads the shoulder all day. [OSHA pointer]
5. Screen at arm's length, with the top no higher than your eyes. Aim for 50 to 100 cm away, with the centre of the screen 15 to 20 degrees below eye level. A stack of books makes a fine monitor riser. On a laptop, raise the laptop and add a separate keyboard and mouse. [OSHA monitors; CCOHS monitor; Asundi 2010]
6. Make the text bigger instead of leaning in. If you catch yourself leaning toward the screen, increase the zoom. [CCOHS monitor]

Then build two habits.

• Get up every 30 minutes or so, ideally for a few minutes' walk. In a small trial, 5 minutes' walking every 30 minutes was the only schedule that lowered blood sugar over a working day, though every break schedule tested lowered blood pressure. [Duran 2023] The trial was small and the breaks were walks, so a walk is the version with evidence. It costs nothing.
• Build up daily activity. In a pooled analysis of about a million people, 60 to 75 minutes a day of moderate activity (self-reported) removed the extra mortality linked to long sitting. [Ekelund 2016] With activity measured by wearable sensors, about 30 to 40 minutes a day of moderate-to-vigorous activity reduced it. [Ekelund 2020] These are observational studies, but large ones: this is the better-supported of the two habits.

What to watch out for

These are the signs that something needs more than an adjustment.

• Numbness, tingling or pins and needles in the fingers, especially at night, or a weaker grip. These are nerve symptoms. See a clinician rather than buying equipment.
• Pain that is still there the next morning, wakes you, or is spreading. Ordinary stiffness eases when you move. Persistent pain needs assessment.
• Headaches, blurred vision or burning eyes late in the day. Often an out-of-date prescription, a screen that is too close or too bright, or simply not blinking. Book an eye test before buying a new monitor.
• Standing all day after buying a standing desk. Lab studies find low-back discomfort building after about 40 to 70 minutes of standing still, and jobs that are mostly standing carry their own risks for the legs and heart. Alternate. [Coenen 2017; Tüchsen 2005; Smith 2018]
• Products sold as cures. No keyboard, mouse or chair on its own has strong trial evidence that it prevents pain. Fit and habits matter more than price. [Hoe 2018 Cochrane]
• Blue-light glasses. A 2023 Cochrane review of 17 trials found they may make no difference to eye strain in the short term (low-certainty evidence). [Singh 2023 Cochrane]
• The laptop on the sofa. Every laptop posture measured bends the neck and wrists. A laptop on the lap adds about 6 degrees of head tilt and 8 degrees of wrist extension compared with a desk. [Asundi 2010]

The calculator

Enter your height, or three quick measurements, then check the desk you have now. The numbers are starting points to adjust to comfort.

How strong is the evidence? An honest summary

The set up, part by part

Chair

Look for five adjustments: seat height, seat depth, a lumbar support you can move up and down, a backrest that reclines at least 15 degrees, and armrests that go low enough to stay out of the way. [OSHA chairs] The HFES workstation standard's draft revision asks for a seat that adjusts over at least 11.4 cm, a lumbar support centred 15 to 25 cm above the compressed seat, and armrests 19 to 29 cm above the seat. [HFES 100 draft §5.19] The WELL building standard asks for chairs that adjust between about 38 and 50 cm high. [WELL v2 Feature 73]

• Seat height: feet flat, thighs about level, knees no higher than the hips.
• Seat depth: the back of the knee clear of the seat edge [OSHA]; two to four fingers is a common rule of thumb.
• Recline: sitting slightly reclined, about 100 to 110 degrees, is fine and often more comfortable than bolt upright. The HFES draft allows 90 to 120 degrees.
• Armrests: low enough that your shoulders do not shrug. Remove them if they stop you getting close to the desk.

Desk and keyboard

• Elbows about level with the keyboard [OSHA], or slightly above. An elbow angle of 90 to 110 degrees is the usual guidance [CCOHS], but a more open angle is fine: the HFES draft allows 70 to 135 degrees [§5.4.2], and in a prospective study an angle over 121 degrees was linked to fewer neck and shoulder problems. [Marcus 2002]
• Wrists straight: bent up or down by less than 30 degrees [HFES draft §5.4.2], and kept close to straight side to side. Bending the wrist towards the thumb by more than 5 degrees on the mouse was linked to more hand and arm problems. [Marcus 2002]
• Keyboard flat rather than tilted up on its feet. [OSHA] A keyboard tray sloping away from you kept more wrist movement in the neutral zone in a small field trial. [Hedge 1999]
• Leave some desk in front of the keyboard to rest the forearms. In a prospective study of new computer users, a keyboard more than 12 cm from the desk edge was protective for the hand and arm. [Marcus 2002] Supporting the forearms is worth doing: one trial of a wide forearm board roughly halved neck and shoulder disorders [Rempel 2006], and in the Cochrane review arm support helped when combined with an alternative mouse, though not clearly on its own. [Hoe 2018]
• A fixed desk too high for you: raise the chair and add a footrest, or fit a keyboard tray that adjusts in height and tilt and holds the mouse too. [OSHA]

Mouse and pointing devices

• Right beside the keyboard, at the same height.
• Use keyboard shortcuts for repeated commands. Every mouse trip you skip is one you do not make ten thousand times.
• Heavy mouse use (over 20 hours a week) was weakly associated with carpal tunnel symptoms in the largest study, and more mouse time predicted short-term neck and shoulder pain, but not long-term pain. [Andersen 2003; Andersen 2008] Mixing input devices, or switching mousing hands for part of the day, spreads that load. That is a reasonable inference from the evidence, not a trial result.
• Vertical mice, trackballs and roller bars change forearm posture in the lab. A vertical mouse is a reasonable thing to try if your forearm aches, but no outcome trial shows it prevents injury.

Monitor

• Distance: 50 to 100 cm from eye to the centre of the screen, measured along your line of sight. [OSHA; HFES draft §4.2] Most people in HFES studies preferred 75 to 83 cm, provided the text was large enough to read there. A very wide screen may need to sit further back so you are not turning your head to its edges; that is our suggestion, not a standard.
• Height: the top of the picture at or below eye level, and the centre 15 to 20 degrees below. A monitor slightly too low is better than one too high. [OSHA; CCOHS]
• Tilt: back 10 to 20 degrees, facing your eyes. Watch for ceiling-light glare when you tilt.
• Text: the HFES draft measures text by its lower-case x height: at least 16 minutes of arc at your eye, and 22 for comfortable reading. [HFES draft §4.9.1] At 75 cm that is an x about 3.5 mm tall, and 4.8 mm for comfort. If you find yourself leaning in, your text is probably smaller than that.
• Progressive or bifocal lenses: lower the screen further, as low as the desk allows (the HFES draft allows 15 to 40 degrees below eye level for these lenses), or ask your optometrist about single-vision computer glasses. [OSHA; HFES draft §4.4]

Two or more monitors

• One main screen: centre it in front of you, put the second beside it angled in, and keep both at the same height. [CCOHS]
• Two used equally: put the seam on your centre line and angle both in, so they sit on an arc at the same distance. [CCOHS]
• Height: with two or more screens the HFES draft puts the screen centre 5 to 10 degrees below eye level, a little higher than for one screen. Keep the tops no higher than your eyes. [HFES draft §4.4; OSHA]
• How far to the side: OSHA says a monitor should be no more than 35 degrees to the left or right. [OSHA monitors] With two 27-inch screens side by side at 75 cm, the second one's centre sits more than 40 degrees out (our arithmetic), so treat it as a reference screen.
• For long stints on a side screen, turn your chair rather than your neck.

Standing desks

A standing desk is a tool for changing position, not for standing all day.

• Set the standing height by the same rule: elbows about level with the keyboard. If the screen stands on the desk or a desk-mounted arm, it rises with the desk and usually needs no separate change. Check the top is still no higher than your eyes.
• How long to stand: an expert panel suggested working up to 2, then 4, hours a day of standing or light activity spread across the working day. [Buckley 2015] That is consensus, not a trial result. Lab data suggest changing position at least every 40 minutes or so. [Coenen 2017] There is no evidence behind any precise sit-stand ratio.
• Wear supportive shoes, or use an anti-fatigue mat.
• An under-desk walking pad or treadmill is a reasonable option if you like it. The Cochrane review found the effect of treadmill desks on sitting time unclear. [Shrestha 2018]

Light and daylight

• About 500 lux on the desk for reading and writing is the European office standard. [EN 12464-1 via DIAL] Screen work is often comfortable in less: the CCOHS range for computer tasks is 75 to 300 lux. Match screen brightness to the room.
• Put windows to your side, not behind the screen or behind you, to avoid glare and reflections.
• Daylight matters beyond the desk. In a small study, office workers with windows slept longer and rated their sleep better than those without. [Boubekri 2014] A walk outside in the morning covers both this and the movement advice.

Eyes

• Blinking slows sharply at a screen, which dries the eyes. One study measured a five-fold drop [Patel 1991]; the AAO gives about 15 blinks a minute normally and 5 to 7 at a screen. [AAO]
• Every 20 minutes or so, look at something across the room for 20 seconds and blink fully. It helps while you do it. The benefit does not last if you stop. [Talens-Estarelles 2023]
• Estimates suggest half or more of screen workers have digital eye strain, though studies define it differently. [Sheppard & Wolffsohn 2018] An up-to-date prescription is the first fix.

Occupational hazards of long-term desk and screen work

Noise, mental workload and circadian lighting are real hazards in desk work, but this draft has not yet gathered sources for them. They are marked for the next pass.

Job analysis: programmers, AI and data engineers, and designers

These figures come from O*NET OnLine (U.S. Department of Labor, data as displayed 2026-11-08) and the BLS. O*NET has no separate title for AI or machine-learning engineers. Data Scientists (15-2051.00) and Computer and Information Research Scientists (15-1221.00) are the closest.

The demand profile

O*NET publishes only the response categories that meet its precision standard, so some distributions are incomplete and "not shown" means unknown, not zero. Static strength is rated 0 in all five. Noise, poor lighting, bending and cramped space are reported as "never" by most respondents in every one.

What each job loads, and what to do about it

Programmers and software developers. The heaviest sitting exposure in the set: 95% sit continually, and 61% work more than 100 feet from colleagues. The load is static, on the neck and lower back, rather than repetitive effort. Surveys of software professionals put 12-month musculoskeletal symptoms at about 69 to 72%, mostly low back and neck, and in one study working 8 or more hours a day roughly tripled the odds. These are small surveys in India (Chaturvedi 2025 surveyed 200 people; Saleem 2015 surveyed software professionals in Chennai), so treat the figures as indicative. [Chaturvedi 2025; Saleem 2015] Priorities: chair and lumbar support, screen height, and getting up. A short walk every half hour is the habit with trial evidence behind it. [Duran 2023] Developers who read code for long stretches can enlarge the text rather than lean in.

AI and data engineers, data scientists. Similar sitting load, but more meetings and calls (43% face-to-face daily for data scientists, 64% telephone daily for research scientists) and the strongest cognitive demand ratings in the set (deductive and inductive reasoning 78 to 81). Long training runs and analysis invite sitting through without a break. Priorities: the same set up as developers, walking calls and standing meetings, and two screens arranged around one primary rather than evenly, since much of the work is one main window with reference material beside it.

Graphic designers. The heaviest hand and eye load: 53% make repetitive motions continually, colour discrimination matters (56), and 70% face strict deadlines daily with 55% working over 40 hours. Precise pointing and drawing are part of the job. Priorities: forearm support, a pointing device close in, and colour-accurate lighting: a neutral, glare-free room with screen brightness matched to it. Pen tablets lowered forearm muscle activity against a mouse in a lab study, but comfort was mixed, so try before committing. [Kotani 2003; Hemati 2020] Place a tablet directly in front of you with the screen above it, not off to one side. Perceived muscle tension and low task variety were among the risk factors for neck pain in office workers [Jun 2017]; whether deadline pressure adds to them was not studied there, so treat that as our inference.

Web and digital interface designers. Between the two: continuous sitting (69%), frequent deadlines (53% daily), and design plus code work that often spans two or three screens. Priorities: monitor arrangement and keyboard-mouse switching.

What the injury statistics do and do not show

The BLS counts injuries and illnesses that cost work days. For 2023 to 2024 the rate of days-away cases for software developers was 1.3 per 10,000 full-time workers, against 86.6 for all occupations. [BLS SOII Table R100] That reflects how rarely desk symptoms become time-off injuries, not how common they are. Smaller surveys abroad find musculoskeletal symptoms in about 70% of software professionals over a year [Chaturvedi 2025, India, 200 people], and digital eye strain in 82% of IT staff at one university [Zayed 2021, Egypt, 108 people]. Across 45 studies of computer users generally, the pooled figure is 66%. [Anbesu 2023] None of these samples matches the US workforce the BLS counts. Desk work rarely causes acute injuries. It is a slow, common, low-grade burden, and that is why setup and habits are worth getting right early.

Graphic designers are an evidence gap: no peer-reviewed prevalence study specific to them could be opened and verified for this guide.

References

1. OSHA. Computer Workstations eTool: Chairs; Keyboards; Monitors; Pointer/Mouse. https://www.osha.gov/etools/computer-workstations
2. CCOHS. Office Ergonomics: Positioning the Monitor; Keyboard Selection and Use; Sit/Stand Desk; Lighting Ergonomics. https://www.ccohs.ca/oshanswers/ergonomics/
3. HFES. BSR/HFES 100-2026, Human Factors Engineering of Computer Workstations, public review draft. https://www.hfes.org/Portals/0/Uploads/BSR%20HFES%20100-2026%20Draft%20%28003%29.pdf
4. IWBI. WELL v2 Feature 73, Ergonomics. https://standard.wellcertified.com/v2/node/12153
5. DIAL. Standard values for office interiors (EN 12464-1). https://dialux4.support-en.dial.de/support/solutions/articles/9000080374-standard-values-for-office-interiors
6. American Academy of Ophthalmology, 2024. Computers, Digital Devices, and Eye Strain. https://www.aao.org/eye-health/tips-prevention/computer-usage
7. Gordon CC et al. 2014. 2012 Anthropometric Survey of U.S. Army Personnel (ANSUR II). NATICK/TR-15/007. https://www.openlab.psu.edu/ansur2/
8. Shrestha N et al. 2018. Workplace interventions for reducing sitting at work. Cochrane CD010912. https://doi.org/10.1002/14651858.CD010912.pub5
9. Hoe VC et al. 2018. Ergonomic interventions for preventing work-related MSDs of the upper limb and neck among office workers. Cochrane CD008570. https://doi.org/10.1002/14651858.CD008570.pub3
10. Singh S et al. 2023. Blue-light filtering spectacle lenses. Cochrane CD013244. https://doi.org/10.1002/14651858.CD013244.pub2
11. Andersen JH et al. 2003. Computer use and carpal tunnel syndrome: a 1-year follow-up study. JAMA 289:2963. https://doi.org/10.1001/jama.289.22.2963
12. Thomsen JF, Gerr F, Atroshi I. 2008. Carpal tunnel syndrome and the use of computer mouse and keyboard. BMC Musculoskelet Disord 9:134. https://doi.org/10.1186/1471-2474-9-134
13. Andersen JH et al. 2008. Computer mouse use predicts acute pain but not prolonged or chronic pain in the neck and shoulder. Occup Environ Med 65:126. https://doi.org/10.1136/oem.2007.033506
14. Hedge A, Morimoto S, McCrobie D. 1999. Effects of keyboard tray geometry on upper body posture and comfort. Ergonomics 42:1333. https://doi.org/10.1080/001401399184983
15. Tittiranonda P et al. 1999. Effect of four computer keyboards in computer users with upper extremity MSDs. Am J Ind Med 35:647.
16. Rempel DM et al. 2006. A randomised controlled trial evaluating two workstation interventions. Occup Environ Med 63:300. https://doi.org/10.1136/oem.2005.022285
17. Gerr F et al. 2002. A prospective study of computer users: I. Am J Ind Med 41:221. https://doi.org/10.1002/ajim.10066
18. Marcus M et al. 2002. A prospective study of computer users: II. Postural risk factors. Am J Ind Med 41:236. https://doi.org/10.1002/ajim.10067
19. Jun D et al. 2017. Physical risk factors for developing non-specific neck pain in office workers. Int Arch Occup Environ Health 90:373. https://doi.org/10.1007/s00420-017-1205-3
20. Asundi K et al. 2010. Notebook computer use on a desk, lap and lap support. Ergonomics 53:74. https://doi.org/10.1080/00140130903389043
21. Buckley JP et al. 2015. The sedentary office: an expert statement. Br J Sports Med 49:1357. https://doi.org/10.1136/bjsports-2015-094618
22. Coenen P et al. 2017. Associations of prolonged standing with musculoskeletal symptoms. Gait Posture 58:310. https://doi.org/10.1016/j.gaitpost.2017.08.024
23. Tüchsen F et al. 2005. Prolonged standing at work and hospitalisation due to varicose veins. Occup Environ Med 62:847. https://doi.org/10.1136/oem.2005.020537
24. Smith P et al. 2018. Occupational standing and sitting and incident heart disease. Am J Epidemiol 187:27. https://doi.org/10.1093/aje/kwx298
25. Ekelund U et al. 2016. Does physical activity attenuate, or even eliminate, the detrimental association of sitting time with mortality? Lancet 388:1302. https://doi.org/10.1016/S0140-6736(16)30370-1
26. Ekelund U et al. 2020. Joint associations of accelerometer-measured physical activity and sedentary time with all-cause mortality. Br J Sports Med 54:1499. https://doi.org/10.1136/bjsports-2020-103270
27. Patterson R et al. 2018. Sedentary behaviour and risk of all-cause, cardiovascular and cancer mortality, and incident type 2 diabetes. Eur J Epidemiol 33:811. https://doi.org/10.1007/s10654-018-0380-1
28. Duran AT et al. 2023. Breaking up prolonged sitting to improve cardiometabolic risk. Med Sci Sports Exerc 55:847. https://doi.org/10.1249/MSS.0000000000003109
29. Albulescu P et al. 2022. "Give me a break!" A systematic review and meta-analysis on micro-breaks. PLoS ONE 17:e0272460. https://doi.org/10.1371/journal.pone.0272460
30. Sheppard AL, Wolffsohn JS. 2018. Digital eye strain: prevalence, measurement and amelioration. BMJ Open Ophthalmol 3:e000146. https://doi.org/10.1136/bmjophth-2018-000146
31. Patel S et al. 1991. Effect of visual display unit use on blink rate and tear stability. Optom Vis Sci 68:888.
32. Talens-Estarelles C et al. 2023. The effects of breaks on digital eye strain, dry eye and binocular vision: testing the 20-20-20 rule. Cont Lens Anterior Eye 46:101744. https://doi.org/10.1016/j.clae.2022.101744
33. Boubekri M et al. 2014. Impact of windows and daylight exposure on overall health and sleep quality of office workers. J Clin Sleep Med 10:603. https://doi.org/10.5664/jcsm.3780
34. Moretti A et al. 2020. Characterization of home working population during COVID-19 emergency. Int J Environ Res Public Health 17:6284. https://doi.org/10.3390/ijerph17176284
35. Gerding T et al. 2021. An assessment of ergonomic issues in the home offices of university employees. Work 68:981. https://doi.org/10.3233/WOR-205294

1. O*NET OnLine, U.S. Department of Labor. Occupation details 15-1252.00, 15-2051.00, 15-1221.00, 27-1024.00, 15-1255.00. https://www.onetonline.org/link/details/15-1252.00
2. U.S. Bureau of Labor Statistics. Survey of Occupational Injuries and Illnesses, 2023-2024, Table R100. https://www.bls.gov/iif/nonfatal-injuries-and-illnesses-tables/soii-case-and-demographic-characteristics-historical-data/2023-2024.htm
3. Chaturvedi R et al. 2025. J Family Med Prim Care 14:2847. https://pmc.ncbi.nlm.nih.gov/articles/PMC12349777
4. Saleem M et al. 2015. Int J Community Med Public Health 2:367. https://doi.org/10.18203/2394-6040.ijcmph20150941
5. Zayed HAM et al. 2021. Environ Sci Pollut Res 28:25187. https://doi.org/10.1007/s11356-021-12454-3
6. Anbesu EW, Lema AK. 2023. Prevalence of computer vision syndrome: a systematic review and meta-analysis. Sci Rep 13:1801. https://doi.org/10.1038/s41598-023-28750-6
7. Kotani K, Horii K. 2003. J Physiol Anthropol Appl Human Sci 22:89. https://doi.org/10.2114/jpa.22.89
8. Hemati K et al. 2020. Work 65:701. https://doi.org/10.3233/WOR-203124

This page is general guidance for setting up a workstation, not medical advice.