The Correct Position to Sit at a Desk Without Wrecking Your Back

Like our Content? Stay connected via Google

Follow us for the latest health and wellness information.

I have spent over thirty years around bodies that hurt, including my own. As a runner, golfer, yoga teacher, trainer and now as someone who writes at a desk for a living, I've spent plenty of time helping my students, clients and myself find ways to feel less pain. Here is what I keep finding: almost every guide to proper desk posture hands you the same set of angles, but almost nobody who reads them is still sitting that way an hour later.

So this guide does two things. It gives you the correct position to sit at a desk and the ergonomic desk setup that makes that position possible to hold, because you cannot sit well at a workstation that fights you. Then it tells you the part the diagrams leave out, which is that the angles are the smaller half of the answer and the research is fairly clear about what the bigger half is.

Disclaimer: Links may contain affiliate links, which means we may get paid a commission at no additional cost to you if you purchase through this page. Read our full disclosure here.

The correct position to sit at a desk

Start with the targets, because you do need a baseline to begin before doing some fine tuning at the end. Work from the screen down, since screen height drives your neck and your neck drags everything else with it.

That is how to sit at a desk with good posture, in one table. Hold it loosely though, because the whole second half of this article is about why holding any shape for eight hours is the real problem.

Your ergonomic desk setup, from the screen down

Everything in that table is about your body. This part is about your desk, and it matters more than people expect, because the fastest way to sit badly is to own a workstation that makes sitting well impossible. Most people do not have a discipline problem. They have a desk that is the wrong height and a screen that is too low.

Start with screen height. Top of the monitor at or just below eye level, roughly an arm's length away. This is the one I would fix first in any setup because a low screen pulls your head forward, your neck holds that weight all day and your upper back rounds to follow. If you work on a laptop, understand the trap: the screen and the keyboard are bolted together, so you can have your neck right or your hands right, never both. A riser plus a separate keyboard is the only real fix and it is the highest value change you can make to a laptop setup.

Then keyboard height, which is where most desks fail. A standard desk is one fixed height built around a taller than average body. If your desk has a pull-out tray, use it, because a tray sits several inches below the desk surface and that gap is usually the difference between elbows at 90 degrees and elbows riding up around your ribs. If your desk has no tray, an under-desk keyboard tray is the cheapest meaningful upgrade in this whole article.

Then your feet, and this is the step almost everybody skips. If you are shorter than about 5'4″, pay attention here. Because the desk height is fixed for a taller default body, raising your chair until your elbows are right usually leaves your feet hanging. Leave them dangling and your low back pays for it, so get them supported even if that means a stack of books under your feet while you shop for a footrest.

If you run two monitors, put the one you use most directly in front of you and angle the second one in toward you. The most common dual monitor mistake is centering yourself on the gap between the screens, which means you spend the whole day turned a few degrees to the side without noticing. Your body does notice.

An ergonomic desk setup for neck pain starts and usually ends with screen height, and it is worth measuring rather than eyeballing. Sit normally and look straight ahead. Your gaze should land at or near the top of the screen. If it lands in the middle, the monitor is too high and needs to come down. If it lands above the screen entirely, it needs to come up.

That is more specific than the advice you usually see, and the specificity matters. OSHA puts the top of the monitor at or slightly below eye level and the center of the screen 15 to 20 degrees below horizontal, at a viewing distance of 20 to 40 inches. “At eye level” is the ceiling of that range, not the target, and a screen any higher than that is the thing quietly extending your neck all day.

Laptop users and anyone in progressive lenses should read that twice. Both setups quietly train you to tip your head back or drop your chin, and OSHA's specific advice for bifocal and progressive wearers is to lower the monitor further than the standard recommendation, because you are otherwise reading through the bottom of the lens with your chin in the air. That was my problem for years and it is why the glasses fix further down this page mattered as much as the monitor riser did.

An ergonomic desk setup for carpal tunnel or any wrist and forearm trouble follows the same logic, and the fix is almost never a brace. It is getting the keyboard low enough and close enough that your wrists stay flat rather than bent back, which usually means a tray below desk height rather than a keyboard sitting on top of the desk. A flatter input device helps too, since a tall contoured mouse holds your wrist in extension for hours. This is the row most people skim past in the table above and it is the one that got me, which I will come back to further down.

Bad desk posture symptoms and what each one is telling you

Almost nobody goes looking for a desk setup guide until something hurts. The useful thing about these complaints is that they are specific, so you can usually work backward from the symptom to the thing that caused it.

• Aching between your shoulder blades. Usually reaching. Your keyboard or mouse is too far forward and your upper back is holding your arms out there all day.
• Neck ache, or a headache that starts at the base of your skull. Screen too low, or a chin tipped back to read through progressive lenses.
• Tingling, numbness or pins and needles in your fingers. Wrists bent back rather than flat, which usually means the keyboard is too high or sitting on the desk surface instead of on a tray.
• Numb, hot or heavy thighs. The front edge of the seat is pressing into the back of your legs. Either the seat is too deep for you or your feet are not supported.
• Shoulders creeping up toward your ears. Armrests or desk too high. This one is almost never a posture habit, it is a measurement.
• Low back ache that arrives two or three hours in. This is the one people misread. It is usually not slouching, it is stillness, and the rest of this article is mostly about that.

The part most desk posture advice gets wrong

There are two camps online and they appear to contradict each other. The ergonomics camp gives you the table above and implies that holding those angles is the whole job. The newer physio camp says posture barely matters and tells you to stop worrying about sitting up straight. You will see that camp repeat a line about the best posture being your next posture. What they mean is that no position is good enough to hold for hours, including the one in that table, so the thing worth working on is not finding the perfect shape, rather it is changing shape often.

Both camps have real data and the reason they seem to disagree is that they are measuring different things.

Posture does show up in the numbers. A 2024 survey of 634 frequent computer users published in the Journal of Occupational and Environmental Medicine found that people who sat with the trunk a little bent forward had roughly 2.9 times the odds of low back pain compared with people who sat straight up with their back against the backrest. Interestingly, whether they worked from home or at an office made no difference at all. The position did.

But look at what happens when you widen the lens to include more studies. A 2025 review in the Journal of Back and Musculoskeletal Rehabilitation pulled together 22 studies covering 7,814 office workers and sorted them by which of three things each study had measured. Of the studies that measured total sitting time, 8 of 13 found a link to low back pain. Of those that measured sitting posture, 7 of 10 did. Of those that measured sitting behavior, meaning how static someone stayed and how often they broke it up, 12 of 13 did.

Line those three up and the pattern is hard to miss. How much these people moved while they sat tracked with back pain more consistently than either how long they sat or how well they held themselves and that is what the review authors concluded. Not that posture is irrelevant, but that of the three things you could measure about a person in a chair, stillness was the one that kept showing up.

Your body tells you before your back does

Researchers in Thailand sat 40 healthy people at a computer and had them type continuously for four hours, rating discomfort in ten body regions along the way. Published in Applied Ergonomics, the results were unambiguous: discomfort climbed continuously in every single region for four straight hours, with the worst of it in the low back, buttocks, upper back, thigh and neck. These were healthy volunteers with no complaint going in.

The detail I find most useful is what those people did without being asked. They shifted position roughly 15 to 18 times an hour and the bigger shifts increased as the hours went on. Your nervous system already runs this program. Fidgeting is not a discipline failure, it is a correction.

So what happens to the people who do not shift? Swiss researchers put pressure mats under 64 call center employees and collected 400 hours of sitting data, reported in Applied Ergonomics. Seventy five percent of them reported back pain of some kind and the people with chronic low back pain trended toward more static sitting than their pain free colleagues. Stillness, not slouching, was the pattern that tracked with the chronic group.

A separate field study of 28 office workers in Applied Ergonomics split people into breakers, who stood up at least once during a two hour block of real work, and prolongers, who did not. The breakers kept moving more consistently while seated and reported lower discomfort. Just one stand up changed how they sat for the rest of the time being recorded.

The movement break dose that has data behind it

Everything up to here has been about how sitting feels. The stiffness, the ache that arrives around hour three, the shifting you do without noticing. That is what sends most people looking for a posture guide in the first place.

But the research on breaking up sitting mostly did not come out of back clinics. It came out of metabolic labs and those researchers were not asking anyone to rate their discomfort. They were drawing blood. What they wanted to know was what happens to blood sugar, insulin and blood pressure when a body stays still for eight hours at a stretch.

That is good news for us, because it means one small habit has been tested against two separate problems and it holds up for both. The break that keeps your back from locking up is the same break that blunts what your blood sugar does after lunch. If blood sugar and blood pressure are already on your radar, that second reason may end up mattering more to you than the first. One behavior, two payoffs, nothing extra to remember.

Which brings us to the numbers. “Get up and move more” is useless advice because it has no number attached. There are numbers though and they come from controlled crossover trials where people sat all day in a lab while researchers drew blood.

The original is still one of the cleanest. In Diabetes Care, 19 overweight adults aged 45 to 65 sat for five hours after a test meal in three conditions. Uninterrupted sitting, or sitting broken by two minutes of light walking every 20 minutes, or the same with moderate walking. Both walking conditions cut the glucose response meaningfully, from 6.9 down to 5.2 with light walking, and dropped insulin from 828.6 to 633.6. Two minutes. Light walking. Every twenty minutes.

A later trial in Medicine and Science in Sports and Exercise went further and tested four different doses head to head over eight hour days.

Read that table as a menu rather than a verdict. If blood pressure is what you care about, even one minute an hour bought a real reduction. If blood sugar is the target, you need the bigger dose of five minutes every half hour. And if back pain is your main concern, choosing any of these movement patterns will help.

These recommendations don't only apply to your workday but rather any time you're sitting for an extended period. A crossover study of 30 adults in Medicine and Science in Sports and Exercise used three minutes of simple resistance moves, things like chair squats and calf raises, every 30 minutes through an evening of sitting. Glucose response dropped 31.5 percent and insulin 26.6 percent against uninterrupted sitting and the effect held across every body mass index category in the study. If you are already strength training that's great but it can't counteract all the sitting. Moving often anytime you sit is still needed no matter how much you exercise.

One caveat, a study of much shorter bouts, 15 to 30 seconds of stair climbing once an hour, published in Medicine and Science in Sports and Exercise, found no significant metabolic benefit in healthy weight young men, though it did lower insulin by 16.5 percent in adults carrying extra weight. So short and occasional may not be enough if you are already metabolically healthy. Two minutes beats fifteen seconds so you may just want to keep your movement period slightly longer just to be on the safe side.

The version you can do without leaving your desk

Every dose above still assumes you can get up and leave your desk and some days you just may not be able to. However, there is a version of this that never takes you more than an arm's length from your chair and it has a name in the research: activity snacks.

Researchers at the University of Toronto put 14 adults through three separate 7.5 hour lab days and published the result in the Journal of Applied Physiology. The conditions were uninterrupted sitting, two minutes of treadmill walking every 30 minutes, or 15 chair stands with a calf raise every 30 minutes. Those chair stands were standardized to 15 in a minute and are essentially a body weight squat. In the hour after lunch, insulin was 37 percent lower with the chair stands and 29 percent lower with the walking, both measured against sitting still. Out at three hours, only the chair stands were still ahead, 24 percent below the sitting condition. One minute of standing up and sitting back down with a calf raise matched two minutes on a treadmill, then outlasted it.

The same lab went back and asked what those squats do to muscle, which is the part I find most interesting. In a follow-up study in the Journal of Applied Physiology, 12 adults ran the same three conditions while researchers tracked labeled amino acids out of the meals and into muscle. Bodyweight squats every 30 minutes and short walks every 30 minutes both raised myofibrillar protein synthesis compared with sitting through the day. Same lunch, same protein, more of it ending up in muscle. The only variable was whether they got up.

After reading this study and now writing about it, I'm working this into my day.

That is the part I don't want you to skim. If you are eating for muscle, which most of you reading this are, then sitting still between meals is quietly working against the meal you just ate. Squats at your desk are not only a blood sugar intervention, they change what your body does with your protein.

Two caveats and then I will let it go. Both trials were small, 14 people in one and 12 in the other and the participants were healthy adults in their twenties rather than women my age, so I am not going to promise you those exact percentages. None were run for months at a time, so all we know is that it may work on blood sugar for sure short term but we don't know what it does to muscle mass over a year.

However, what I take from it is directional and it is enough to act on. Chair squats at your desk are not the consolation prize for a day you could not get out for a walk. In the one head to head that measured both, they held up longer than the walk did.

I wrote a short piece on movement snacks after sitting too long as part of the 30 Day Healthy Habits Challenge, which is the same idea under a friendlier name than the researchers gave it.

Desk posture exercises worth your time

A word on what these are and are not. The movement break research above is about interrupting sitting, and none of those trials tested the specific stretches below. These are what I give clients and what I do myself as a trainer and yoga teacher, aimed at the places a desk day tends to shorten. Think of them as the thing to do when you get up, not as a replacement for getting up.

1. Chin tucks. Sit tall, look straight ahead and draw your chin straight back like you are making a double chin, without tipping your head down. Hold five seconds, do ten. This works the deep neck flexors that spend all day being overridden by a forward head position.
2. Doorway chest opener. Forearm on the door frame, elbow at shoulder height, step through gently until you feel it across the front of your shoulder. Thirty seconds each side. Hours of reaching forward shortens the chest, and a tight chest is what pulls your shoulders round.
3. Thoracic extension over the chair back. Sit, lace your hands behind your head, and lean back over the top edge of your chair so the bend happens in your upper back rather than your low back. Five slow reps. This is the one that feels best after a long stretch of typing.
4. Standing hip flexor stretch. Half kneeling or in a split stance, tuck your pelvis under and press the back hip forward. Thirty seconds each side. Sitting keeps your hip flexors in a shortened position for hours and that shows up as low back tightness far more often than people realize.
5. Wrist extensor stretch. Arm straight out, palm down, gently pull your fingers back toward you with the other hand. Twenty seconds each side. Pair it with fixing your keyboard height, because the stretch alone will not undo eight hours of bent wrists.

None of this takes more than three or four minutes and you do not need all five in one go. Pick the one that matches wherever you feel it most, which the symptom list further up should help you identify.

Why women may get more out of breaking up sitting

This is the part I had not seen covered anywhere and one of the reasons I bothered writing this article. Researchers pooled four similarly designed trials, 129 people in total and looked at who responded best to interrupting sitting with light activity. The analysis, in Medicine and Science in Sports and Exercise, found that women got a larger reduction in postprandial insulin than men, 21.2 percent against 17.6 percent. The glucose gap was wider still, a 6.8 percent reduction in women compared with 1.7 percent in men. Having a higher body mass index also predicted a bigger response.

That 2024 survey of computer users I mentioned earlier turned up something I saved for here. Among that same group of computer users, being female carried about 2.2 times the odds of low back pain. Put the two findings side by side and the picture for a woman at a desk is that she is at somewhat higher risk of problems and stands to gain more from the fix.

None of these studies were designed to test perimenopause or menopause, and none of them measured hormones. But, in the trials we have, women got more benefit per movement break than men did. That is enough reason for women especially to take the breaks seriously. If you want the bigger picture on what more serious movement does for a woman in this age range, I covered it in my guide to strength training for women over 40.

Does a standing desk fix this?

Partly and not in the way most people buy one expecting.

In that same pooled analysis of 129 people, interrupting sitting with light activity lowered insulin and glucose. Interrupting it with standing did not. Standing up is a different intervention from moving and the metabolic benefit in that dataset came from the moving.

Where standing desks do earn their place is comfort and total sitting time. A 2024 study in IISE Transactions on Occupational Ergonomics and Human Factors compared administrative workers using traditional desks, sit-stand desks and stand-biased desks, which default to standing height with a tall stool available. The stand-biased group stood more, sat less and reported significantly less lower back discomfort than both the traditional desk group and the sit-stand group. The default position mattered more than the capability, which tracks with everything we know about how people use adjustable equipment in practice.

Stand up desk posture follows the same logic as sitting: screen at eye level, elbows near 90 degrees, weight even across both feet rather than parked on one hip. Shift your weight, use a footrest bar or a low box to rest one foot on and change which foot every so often. And the same rule applies as everywhere else in this article. Standing still for three hours is its own static posture, just a vertical one.

Two pieces of honesty about this whole evidence base. A randomized workplace intervention in the International Journal of Environmental Research and Public Health ran 153 desk workers through three months of sitting reduction and found no substantial average change in pain, though people who cut their sitting the most did see less low back pain, on the order of a 0.84 point drop per hour of sitting removed. And an overview of systematic reviews in Applied Physiology, Nutrition, and Metabolism rated the overall certainty of the sedentary behavior evidence as low to very low, but the direction is consistent.

What your chair can and cannot do

A good chair does three things. It puts your elbows and feet where the table above wants them, it makes contact with your lumbar curve instead of hovering behind it, and it lets you change position without fighting the seat.

What no chair does is make you move. Given that static sitting is the pattern that showed up alongside chronic back pain in the pressure mat data, a chair so comfortable that you never shift is not obviously doing you a favor.

The practical version: if your current chair has a working height adjustment and a backrest you can reach without sliding forward, spend your money on a monitor riser and a timer before you spend it on a new chair. If your feet dangle or your elbows ride up around your ribs, the chair or the desk is the problem and no amount of remembering to sit up straight will fix it.

My own desk setup, including what I got wrong

Several years ago, I had a treadmill desk for a while and I loved it. In fact, that was my setup while writing my book back in 2020. I was able to arrange it pretty easily with an adjustable desk and a walking pad, both from Amazon. When we sold that house and did the Vagabond Tour for a while, that desk and pad got sold as well.

What I use now is unremarkable, my previous older desk that's more traditional. It's a wooden one with a pull-out tray, not a purpose built workstation, and the tray is the piece doing the most work. It sits several inches below the desk surface, which is the way I get my elbows down near 90 degrees without raising the chair until my feet leave the floor. That is the fixed height desk problem from the top of this article, solved for nothing because the desk already had the tray. If yours does not, an under-desk keyboard tray is the cheapest meaningful upgrade in this whole article. This is the one we got for my husband's desk.

On the tray I run a curved keyboard with a built-in palm rest, so my hands sit at an angle rather than twisting out to meet a straight keyboard. The slim Apple keyboard beside it is there for just one reason, the fingerprint reader, so it stays on the tray even though I do all my typing on the curved one. Next to both is a flat trackpad rather than a mouse, which keeps my wrist closer to neutral than the tall contoured mice I used for years. Wrists are the row most people skim past in the table at the top of this page but that's the one that got me. I've had pretty significant wrist and elbow pain over the years during times of heavy desk work. The flat trackpad has helped with that a lot.

The monitor is propped up on a hardback book, with an external hard drive and a dock stacked beside it. It is not pretty, but it puts the top of the screen at eye level. A monitor stand would look better of course, but this works just fine. Height is the fix, the product is optional.

The chair is a mesh backed Herman Miller ergonomic chair I bought off Craigslist years ago, with a seat cushion I added while working around a hamstring injury. The cushion was the piece I resisted longest and it made the biggest difference to how long I can sit comfortably, which is a fair illustration of the general point: the cheap fix aimed at your specific problem usually beats the expensive fix aimed at the category.

I also strength train five days a week and you may wonder if that is enough to counteract sitting most of the day. The trials above put people in chairs for whole days and measured what happened. Being fit did not exempt anyone. If anything, the study on 15 second stair climbing bouts suggests healthy metabolisms respond less to token breaks, so the fitter you are the less you can get away with a fifteen second gesture. You still have to get up.

For more about what lifting does for body composition, read strength training to lose weight.

The habit that stuck for me was attaching the break to something that already happens rather than to a timer I would learn to ignore. I talk a lot about that idea, habit stacking, in the 30 Day Healthy Habits Challenge. Every time I would finish a draft section, I would get up to move and refill my water. After reading the research for this article, I've started integrating chair stands instead, which is the whole reason that section is in this article.

A desk reset you can do in ten minutes

Do these in order, because each one changes the next.

1. Set chair height by your elbows, not your legs. Sit, rest your hands on the keyboard and raise or lower the chair until your elbows sit between 90 and 110 degrees with your shoulders down.
2. Now look at your feet. If they are not flat on the floor, you need a footrest. A stack of books works for a week while you decide.
3. Slide back until you touch the backrest. If you cannot reach the lumbar support without sliding your hips forward, the seat is too deep for you and a cushion behind your low back is the cheap correction.
4. Raise the screen. Top of the monitor at or just below eye level, about an arm's length away. Laptop users, this means a riser and a separate keyboard, there is no other way to get both your neck and your hands right at once. Something that made a big difference here for me was also my glasses. I wear progressive lenses and noticed I was always sitting with my head tilted back to read out of the bottom of them which was causing a lot of pain. Instead, I was able to get a pair of dedicated computer glasses which totally solved that. I also found the solution for glasses that are actually affordable, Zenni. I got my prescription computer glasses for about $40 shipped and you can easily use your existing prescription. There is a formula they use to convert it to computer distance.
5. Pull the keyboard toward you. If you are reaching, your upper back is doing the reaching.
6. Pick your break dose and put it somewhere you will see it. Five minutes every thirty is the strongest version. One minute every hour is the floor, and per the blood pressure data it is far from nothing.
7. Pick your desk-bound fallback too. Fifteen chair stands with a calf raise, about a minute, for the blocks where you cannot walk anywhere.

Then stop trying to hold a perfect desk position. Just set your baseline posture and come back to it when you notice you have drifted and get up on schedule. The evidence points at the getting up being way more important than a proper desk posture anyway!