Does your head really weigh 27 kg when you look down?

13 September 2026 · 6 min read

You have seen the picture. A silhouette bent over a phone, and beside it a number that climbs the further the head tips forward — 12 kg, 18 kg, 27 kg — with a caption comparing it to carrying a small child on your neck. It is one of the most reproduced graphics in the history of posture advice.

It comes from one paper. Kenneth Hansraj published it in Surgical Technology International in 2014, and almost every version of that graphic since traces back to it. It is worth knowing exactly what that paper did, because the answer is more interesting than either the people who repeat the number or the people who dismiss it tend to admit.

What the paper actually computed

Hansraj built a model of the cervical spine in CosmosWorks, a finite element package, assumed a head mass of 13.2 pounds — about 6 kg, which is a reasonable adult figure — and calculated the force at the base of the neck as the head tilted forward through a series of angles.

So the famous 27 kg is the 60° figure, which is roughly where your head sits when you look down at a phone in your lap. And the first thing to notice is that even at 0°, in a perfectly neutral position, the model reports 4.5 to 5.4 kg — noticeably less than the 6 kg head it started with.

That is not an error. It is the clue to what the number means.

It is a moment, not a weight

Your head does not get heavier when you look down. Mass does not change with angle. What changes is the moment arm — the perpendicular distance between the head's centre of mass and the joint it pivots around.

Hold a 5 kg dumbbell against your chest and it is easy. Hold the same dumbbell at arm's length and it is brutal. The weight is identical; the leverage is not. Your neck extensors work against exactly that geometry. Tip the head forward and its centre of mass swings away from the cervical spine, the lever lengthens, and the muscles have to generate proportionally more force to stop your head falling to your chest.

Framed that way, the underlying claim is uncontroversial biomechanics. Flexion multiplies cervical load, and the multiplication is steep — the model puts 60° at roughly five times neutral. Anyone who has held a phone at chest height for an hour has felt the mechanism directly.

Three reasons not to quote the number as fact

Nothing was measured. This is a computed output of a model, not a reading from an instrument in a person's neck. Change the assumed head mass, the assumed centre-of-mass location or the assumed muscle line of action, and every figure in that table moves. The paper reports one set of assumptions, not a range.

The soft tissue is lumped together. The paper states that muscles, tendons and ligaments were incorporated collectively rather than modelled independently. In a real neck those structures share load in ways that shift with fatigue, with co-contraction, and with whether you are supported. A model that treats them as one bundle cannot capture that.

The paper flags its own ceiling. Hansraj notes that at 90° the model's prediction was not reliable. That is honest of him and routinely ignored by everyone reproducing the graphic, some of whom extrapolate past it.

There is also the plain matter of provenance: it is a single-author paper in a journal that is not a biomechanics journal, and it has been repeated far more often than it has been tested. Repetition is not replication.

What this means for your desk

The number is soft. The direction is not. Every extra degree of sustained neck flexion costs you, and the cost rises faster than the angle does — which is why the practical advice survives even when the figure does not.

• Raise the screen, don't lower the head. The top of the display roughly at eye level takes most of the flexion out of the day without you having to think about posture at all.
• A laptop on a desk is the worst case in this respect, because the screen and keyboard are locked together at a height that suits neither. A stand plus a separate keyboard fixes the geometry that no amount of sitting up straight can.
• Duration matters more than angle. Thirty degrees for six hours is a bigger problem than sixty degrees for six minutes, and almost nothing in the popular coverage says so.

That last point is the one I would keep if you kept only one. The graphic implies a threshold — a bad angle to avoid. The mechanics imply an accumulation: load multiplied by time. Which is why the fix is rarely a posture and usually a setup.