How long should you stand at a standing desk?

The honest answer to this question is that nobody knows, and it is worth sitting with that for a second before we get to the numbers.

No source we could find — not Cochrane, not the British Journal of Sports Medicine, not Cornell — establishes that any specific sit:stand ratio is optimal. None of them compares ratios head-to-head. The number you have seen quoted everywhere is real, it has a real author, and it is a rule of thumb on a university web page that its own author explicitly hedges. Everything else is people repeating it with progressively more confidence than the source has.

That is not a reason to ignore it. It is a reason to know what you are holding. So here is where 20-8-2 comes from, what it actually says, and the line inside it that almost nobody repeats — which is, as far as we can tell, the only part that matters.

Where 20-8-2 actually comes from

It comes from the Cornell University Ergonomics Web (CUErgo), the site run by Prof. Alan Hedge. The rule is printed there, in Cornell’s own words:

“20 minutes sitting (in a good posture), 8 minutes standing (for sit-stand workstations) and 2 minutes of standing and moving”

That is a real source. Cornell is a real institution, Hedge is a real and well-published ergonomics researcher, and the page is his. What it is not is a study. There is no trial behind those three numbers, no control group, no published methodology. It is guidance on a web page.

And Cornell says so. The same page calls 20-8-2 “a ball park goal for organizing work” and adds: “These numbers aren’t hard and fast a company can design their work so employees can be more active.” The author is telling you, on the page the rule lives on, that the precision is not there. He is more careful about it than almost anyone who quotes him.

The mis-attribution

Here is the part worth checking, because it is repeated constantly: you will very often see 20-8-2 described as having been published in the British Journal of Sports Medicine. Sometimes it is framed as “researchers writing in the BJSM recommend 20-8-2”, sometimes as a bare citation to a 2015 paper.

The paper being pointed at is the 2015 expert statement “The sedentary office: an expert statement on the growing case for change towards better health and productivity” by Buckley, Hedge, Yates and colleagues, commissioned by Public Health England. We read its full text and searched it.

It contains no “20-8-2”. It contains no sit:stand ratio of any kind.

We are not going to speculate about how deliberate any of this is, and we do not need to — the likeliest explanation is mundane. Alan Hedge, whose Cornell page carries 20-8-2, is a co-author of the BJSM statement. One author, two documents, and over enough repetitions the number drifts from the web page to the journal. But the effect is the same whatever caused it: a rule of thumb ends up wearing a peer-reviewed journal’s authority for a figure that journal never printed. If you have been repeating 20-8-2 because it “came from BJSM”, that specific reason for trusting it does not hold.

What the BJSM statement does say

It is a genuinely useful document, and its actual recommendation is different in kind from a ratio. It is a daily accumulation, and it is graded:

“Initially progress towards accumulating at least 2 h/day of standing and light activity (light walking) during working hours, eventually progressing to a total accumulation of 4 h/day”

Three things about that are easy to miss. First, it is standing and light activitytogether — walking is inside the target, not an extra. Second, it is progressive: 2 hours first, 4 hours eventually, which is a different instruction from a fixed cycle you run all day from day one. Third, and most importantly, the paper grades its own recommendations, and this one is graded (B and C).

That grading scale is the most useful thing in the paper and the least quoted. In the statement’s own terms: (A) means randomised controlled trials with overwhelming data; (B) means RCTs with limited data plus high-quality observational work; (C) means non-randomised or observational evidence; and (D) means panel consensus and expert opinion.

No recommendation in the paper reaches grade A. Several are (D) — which is to say, several are a group of experts stating what they think, clearly labelled as such. That is not a criticism of the paper. The paper is being straight with you about exactly how much weight each line will bear. It is a criticism of everyone who cites it as settled science.

What Cochrane found

The strongest evidence in this area is the Cochrane review “Workplace interventions for reducing sitting at work” (Shrestha N, et al., CD010912), which pooled 34 studies covering 3,397 participants. Its conclusion, verbatim:

“At present there is low-quality evidence that the use of sit-stand desks reduce workplace sitting at short-term and medium-term follow-ups. However, there is no evidence on their effects on sitting over longer follow-up periods.”

The effect size is real and modest. Sit-stand desks reduced sitting by roughly 100 minutes per workdayin the short term (95% confidence interval −116 to −84, across 10 studies, rated low quality), falling to about 57 minutes per day at medium-term follow-up. Note the direction of that drift: the effect shrinks as time passes, and beyond the medium term there is simply no evidence either way.

Read what that outcome actually is, because it is narrower than it looks. Cochrane measured sitting time. Not health. On health, the review’s own plain-language summary calls the benefits “unproven”, and we are not going to write past that. A sit-stand desk demonstrably changes how much of your day you spend sitting, at low quality of evidence, for a while. What that does to you is not established, and this site is not going to be the one that pretends otherwise.

Two more findings from it worth having. On weight, Cochrane is blunt: standing “hardly increases energy expenditure, so one should not expect a sit-stand desk to help in losing weight”. And on the other side of the ledger — a point that deserves more airtime than it gets — the review found no harms. No musculoskeletal pain, no varicose veins, no drop in productivity were reported. The case against these desks is not that they hurt. It is that the case for them is thinner than the marketing.

The line everyone skips

Go back to the Cornell page, because there is one more sentence on it, and it is the one that almost never survives into the blog posts:

“Simply standing is insufficient. Movement is important to get blood circulation through the muscles.”

That is the takeaway, and it quietly demolishes the way the rule is normally used. Look at what happens to 20-8-2 in transmission: the 20 and the 8 survive, because they are the sitting-versus-standing numbers a desk company can sell against. The 2 — the standing and movingminutes, the only part of the cycle that is about movement at all — is the part that gets flattened into “stand for 8” and dropped.

The rule’s own author is on record that the standing is not the mechanism. And the BJSM statement points the same way from the other direction: its target is standing and light activity, and it warns that “Similar to the risks of prolonged static seated positions, so too should prolonged static standing postures be avoided” — graded (D), expert opinion, and we will label it as such. But the word staticis doing all the work in that sentence, and it is the same word doing all the work in Cornell’s.

So the popular version of this rule keeps the two numbers that sell a desk and discards the mechanism its own source says is the point. A desk that goes up and down does not make you move. It removes one excuse not to.

So what should you actually do

We are not going to invent a number to replace the one we have just spent a page qualifying. Here is what the sources support, at the strength they support it.

The direction is supported; the precision is not.Sit less, break it up, and move — that is where Cornell, BJSM and Cochrane all point, at low to moderate quality of evidence. Nothing establishes that 20-8-2 beats 30-10-2, or 45-15, or getting up whenever your coffee runs out. If a page tells you one specific cycle is correct, ask it where the head-to-head comparison is, because we could not find one.

Use 20-8-2 as a prompt, not a prescription.It is a reasonable, well-intentioned rule of thumb from a credible ergonomist, and as a way of remembering to change position it is fine. Treat it the way its author asks you to — a ball park goal — rather than as a dosage.

If you keep only one line from all of this, keep the movement one. Not because we are being contrarian, but because it is the only part of the rule that its own source describes as the mechanism, and it is the part the internet threw away.

And one honest note on where this leaves the purchase. Cochrane found these desks cut sitting time by about 100 minutes a day at low quality of evidence, with the effect shrinking over the medium term, no established health benefit, and no harms. That is a real but unglamorous finding, and it is a perfectly reasonable basis for buying one — if that is the thing you want. It is not a basis for the claims on most of the product pages. If you are not sure the standing half will get used, a convertersits on the desk you already own and answers that question cheaply. And if you are, the thing that actually decides whether you use it is whether it fits you — which is a matter of geometry, and we cover it in the desk height calculator and the standing desk roundup.