I know there are many historical accounts of soldiers trying to cross a body of water and drowning, but this got me thinking how much of this was due to panic, lack of swimming skills, water currents and how much was this due to the weight of their armor and their equipment.

I know for a fact that you can swim in a lorica segmentata, I have a friend that owns a set and managed to swim for a couple of minutes in it, also it turns out that having a scutum on you helps (because it floats...).

So my question is : Given a soldier in peak physical condition who is also a good swimmer, what would be the heaviest set of steel armor that he could wear and still be able to swim and which sets would be just too much?

  • 9
    $\begingroup$ How far do you want your soldiers to swim in it? $\endgroup$
    – Erik
    Jun 17 '19 at 9:17
  • $\begingroup$ Would he care that it will corrode (even quicker through salty water) or does he have a fancy stainless ironman equipment? $\endgroup$
    – Risadinha
    Jun 17 '19 at 10:23
  • 2
    $\begingroup$ Is a submarine considered steel armor? what about a diving tank? a boat? in theory, you can wear as much armor as you have space for. Just add air. $\endgroup$
    – tuskiomi
    Jun 17 '19 at 14:42
  • $\begingroup$ Whilst you ask about steel armour, historically people often did not wear just steel armour (steel plate or mail will stop a blade from cutting but it will not stop the blunt force impact of the strike). A gambeson - padded armour which was typically made of linen - was worn underneath the metal. Linen, being made of flax, would float more easily than steel, thus drastically changing your answer. So then, with this in mind, are you asking how much steel armour (with a gambeson underneath) someone could wear or simply how much steel they can put on their body before they can no longer swim? $\endgroup$ Jun 17 '19 at 18:59
  • $\begingroup$ Do you have some more information on this swim in a Lorica Segmentata? From what I gather a set weighs in at ca. 9kg, which is already a lot more that I would believe someone can swim in... $\endgroup$
    – fgysin
    Jun 18 '19 at 12:10

10 Answers 10


Steel's density of ~7.75 times that of water means you can discount 13% of its weight to buoyancy, but that's about it.

I have lifted 18 kg of collected dropped weightbelts from the bottom on one occasion, which put me at -10 kg of surface buoyancy after my gear weight and wing buoyancy was accounted for, and -14 kg at the bottom for the initial ascent (due to wing compression). This was a major physical effort and I had to hand the belts to the boat ASAP, but I was able to stay sufficiently afloat to ask for the assist.

Generally an experienced technical diver wearing jet fins can sustain about 12 kg of upward thrust, with fins, during an emergency ascent with failed buoyancy devices. Peak static thrust has been measured at 15-19 kg for ~90 kg body weight professional divers. Producing upward force is not equivalent to swimming, as your lateral speed will be very low. It's just a struggle to get to the surface, and you could probably brave a very narrow stream like that.

This is not the average. The average sustained thrust with fins was measured at about 64-69 N, or just 7 kg. Note that this is total prolonged sustained thrust, while practical swimming requires comfortably reaching up for air, but it could be done over a prolonged swim (wide river).

A practice among good swimmers called "monkey diving" involves wearing no BCD (buoyancy compensators) and compensating for buoyancy changes with swimming thrust - so this can be considered a practical swimming weight. The buoyancy at the beginning of a monkey dive can be -3 kg. This takes some effort, but is easily manageable with fins.

Without fins, humans produce very limited static thrust. I can stay afloat and swim with the aforementioned -3 kg of buoyancy without fins, but it's exhausting and slows me down. I can carry more briefly, but -3 kg is as much as I'd be willing to risk carrying across more than 400 meters without the ability to ditch the weight, and -5 kg in a do or die situation (the difference is major: -3 is struggling to get ahead, -5 is struggling to get a breath at all). My weight and swimming fitness would be in the range for the kind of character you describe.

Your average medieval soldier was certainly not a skilled diver, or an skilled swimmer, nor did they have any fins at all. This limits their ability to overcome negative buoyancy to -1 kg for most, and maybe -3 to -6 kg for the best swimmers, with a fairly large body for the era. This number is for swims across calm waters; large lakes, very wide or fast rivers, open sea can be challenging as it is (for that reason, everything above and below is for fresh water).

A sleeveless mail vest weighs about 5 kg. Armor is useless without a weapon (another 1-2 kg), so there's no point in bringing it even if one could. It's well possible to make lighter armor, for instance a steel plate with coverage similar to a SAPI insert, and such plates were sometimes attached to mail or leather. But it's unlikely that someone would bother doing that just for swimming, when it's more practical to supply waxed leather bags for buoyancy instead, or make a raft on the spot.

So the short answer is: Without fins or any buoyancy device - a simple log will do - you can't count on swimming over a decent-sized river with any kind of commonplace medieval steel armor that would be useful on its own.

A good swimmer without fins would still be able to cross rivers with their weapon and their leather armor pieces. They could carry something like steel bracers, but such armor is of limited use without torso protection. A short sleeveless mail vest (not a likely item for a soldier) would be the most a good swimmer could bring.

A shield would be a good flotation aid, and early styles (lime wood with little metal) would be able to support about the shield's own weight in steel. As such aids interfere with swimming, their buoyancy replaces dynamic swimming thrust rather than add to it, but it's a much more practical way of crossing rivers than rushing it.

If we go outside the military, professional swimmers such as pearl divers did exist at the time, and would be more capable. But finding one that turned soldier would be a one in a million occurrence and wouldn't make an army.

To put an upper bound on what's possible, modern Olympic-level athletes produce about the same static thrust without fins as a skilled diver with fins, so that ability to carry 10-12 kg across a river (only just, at risk to their life) could also be expected of them. This level of swimming fitness takes years of training and only came to exist with the reestablishment of full-time professional athletics in the early 20th century.

  • $\begingroup$ Comments are not for extended discussion; this conversation has been moved to chat. $\endgroup$
    – L.Dutch
    Jun 18 '19 at 3:41

A number of people have attempted swimming in heavy medieval-style armor as experiments, with videos available on YouTube or similar sites. I've never seen one where the person actually succeeded. They tend to be at a point where they speculate that with more strength and training they could do better, but... no success so far.

enter image description here

See here for a sample of three such videos.

  • 16
    $\begingroup$ Ahhh... empirical evidence. Yowzah. +1 $\endgroup$ Jun 16 '19 at 1:50
  • 7
    $\begingroup$ Big +1 - nothing answers reality checks more honestly than people checking things in reality :D $\endgroup$
    – Mołot
    Jun 17 '19 at 9:17

Fat warriors could carry more.

We will upfront dismiss assertions that fatness is incompatible with fitness which is of course ridiculous. Next: @Therac's answer above covers upwards thrust produced by arms and legs to counter additional weight.

Let us consider your fat warrior. We will use Andre the Giant as our example. Andre was a 201 kg warrior with 27% body fat. That is 54kg of fat. Fat is 90% of the density of (fresh) water and so 1 kg of fat + 100g extra mass will be neutrally buoyant.

In other words every kg of fat your warrior will, in addition to itself, float 100g of additional mass or 10% of its own mass. 10% of 54 kg is 5.4 kg and so by virtue of the extra fat Andre could float with extra weighty metal gear and compared to some lean marathon running warrior. This is separate from any effort expended kicking to stay afloat; Andre would just bob around.

You can scale that up as far as you want although I expect diminishing return for a standard human frame at very high amounts of fat weight. These doughty warriors will of course float even better in salt water than in fresh by virtue of the higher density of salt water, but their gear would get rusty faster.

  • 16
    $\begingroup$ My only concern here is that it takes more metal to cover a fat warrior than a slender one, which might compensate for the added buoyancy of the fat. Nevertheless, I consider myself unworthy to read this answer - I wouldn't have thought of it in a million years.... +1. $\endgroup$ Jun 15 '19 at 19:07
  • 9
    $\begingroup$ @JBH - true, true <fingers beard thoughtfully>. I think one could graph two lines, one with increased buoyancy from added fat which increases as the cube and one with coverage needs which increases as the square. The spherical ideal warrior will have the minimum surface coverage needs per unit volume, and so would make the most sense for buoyancy applications. $\endgroup$
    – Willk
    Jun 15 '19 at 19:42
  • 22
    $\begingroup$ @Willk are you saying that the best warrior ever is a spherical cow? $\endgroup$ Jun 15 '19 at 21:31
  • 28
    $\begingroup$ @JohnDvorak - yes, and also a good roll model. $\endgroup$
    – Willk
    Jun 15 '19 at 23:24
  • 2
    $\begingroup$ @JBH I had the same thought. Having said that, volume increases faster than surface area (the power of 3 vs. 2), so it's still a net gain. $\endgroup$
    – Sparhawk
    Jun 16 '19 at 22:36

If your soldier knows he will be swimming in armour, you can provide him with an air-filled float. The float can even be made of steel! This can either be a permanent part of his armour, or an attachment.

Some suits of armour already have a bulging breastplate which could house a float (though this may not be the best buoyancy distribution as it could cause him to flip over on his back.

A litre of steel weighs 7.5kg, so for every litre of steel you would need 7.5 litres of air (displacing 7.5kg of water) for neutral buoyancy.

As you add bulk you add more frictional resistance, but this can be compensated by flippers to improve swimming efficiency. And a shield that doubles as a body board / surfboard - why not?

  • 1
    $\begingroup$ The "flip onto his back" might actually be an advantage - if he can manage backstroke or sculling, then it will keep his head out of the water like a life jacket. $\endgroup$ Jun 17 '19 at 7:45
  • 1
    $\begingroup$ Steel armour was always worn over some sort of padding, often multiple layers of linen. This would have trapped air for a short period and provided some buoyancy.. $\endgroup$ Jun 17 '19 at 8:52
  • 1
    $\begingroup$ Many modern soldiers are often carrying an almost-waterproof backpack they can use as a flotation device; look up 'Army Water Survival Training' for pictures. Of course, they still prefer things like rope bridges as not every soldier is a strong swimmer. $\endgroup$
    – mjt
    Jun 17 '19 at 12:47
  • $\begingroup$ @RobinBennett, yeah but once that linen absorbed the water, he'd be pretty darned useless on land. $\endgroup$
    – ShadoCat
    Jun 18 '19 at 22:27

Other answers here seem to looking at the ability to swim under load from a diving perspective, I believe this may not be applicable as many diving techniques rely on flippers.

Water polo players and synchronized swimmers will be familiar with what is called the eggbeater kick, this kick is very effective at providing constant upwards force with reasonable energy demands. This technique is what allows water polo players to launch up out of the water to pass and take shots and for synchronized swimmers to lift their partners out of the water. From a personal perspective this is also what allows for lifeguards to lift and swim with a 9kg (20lb) brick used in training.

This paper suggests that the buoyant force produced by the eggbeater kick may range from between 10-20% of the person's body weight. This would mean that a warrior weighing 180lbs could potentially support themselves in the water carrying 36lbs of amour.

Note: eggbeater kick is primarily effective at providing upwards force and not as much horizontal force, so while a warrior could use this to support themselves in the water they wouldn't be moving quickly. They would also be at much higher risk in choppy water as their head height under such a load would be very close to the surface of the water and waves would cause frequent interruptions to their breathing.


I will go with a 0th order approximation: the human staying still in water, just floating.

The average density of the human body is $985\ \mathrm{kg/m^3}$, and the typical density of seawater is about $1020\ \mathrm{kg/m^3}$. (a) The average density of the human body, after maximum inhalation of air, changes to $945\ \mathrm{kg/m^3}$. (source)

This means that, for a $80\ \mathrm{kg}$ human, his body will occupy after maximum inhalation $0.085\ \mathrm{m^3}$ and will float displacing $0.078\ \mathrm{m^3}$ of seawater. This leaves an additional weight of $(0.085\ \mathrm{m^3}-0.078\ \mathrm{m^3})\times 1020\ \mathrm{kg/m^3}=7.14\ \mathrm{kg}$, which can be added without sinking, if the density of the added material is higher than the density of water.

Considering that one’s nose will sink underwater before the head, and that one cannot be constantly at maximum inhalation, the additional weight is slightly less.

  • 1
    $\begingroup$ "Additional weight that can be added without sinking" isn't the same as "additional weight that can be carried whilst actively trying not to drown", though. $\endgroup$ Jun 15 '19 at 8:36
  • 1
    $\begingroup$ Your calculation dose not take into consideration surface area and movement, as it is right now it only shows what would happen if the human would just be standing still and curled up. $\endgroup$
    – TobyB
    Jun 15 '19 at 8:40
  • $\begingroup$ Swimming with your back down means that the nose is approximately the highest point of your body. $\endgroup$
    – d-b
    Jun 15 '19 at 18:38

If this was a planned swim with time to make something special, replacing much of the leather undeneath (which is an important part of the armour but has at best neutral buoyancy when wet) with a light wood would make things much better. Perhaps a composite formed of balsa on the inside to provide buoyancy and padding, with laminated alder, spruce or pine (less buoyant but stronger) against the metal. The idea would be to form plywood in place; 2D bends are easy but 3D bends are possible. You would of course need a non-soluble glue.

If the wood has an average density of 200 kg/m3 (i.e. mainly balsa) and is used to support 1.5 mm steel with a density of 8000 kg/m3 you'd need 13 mm of wood (1/2") to acheive neutral buoyancy. This seems a little large but not impossible at least for greaves and a cuirass; a stripped down version of the latter (steel breast-plate plus protection for neck and shoulders, with a thick collar of buoyant material to support the head). Strapping a helm and weapons to a shield a towing them would be a good idea.

Bamboo has also been used in the construction of armour (even occasionally on its own) and floats well; armour made almost completely of bamboo would easily float. If the swim was for a sneak attack from a lightly armoured force that would be one approach.

Even if you have netutral buoyancy or better, swimming encumbered is tiring and not quick. I've been known to swim in full kayaking gear; and even a buoyant, open helmet drags on the water, hence avoiding swimming in a heavy helm. A buoyancy aid is thicker than I've suggested for the armour but drags a lot.


I know you asked for steel armor, but Samurai had a special swimming style just for swimming in armor.


Nihon Eiho, or samurai swimming, began as a fundamental martial discipline, a military skill as highly regarded as horsemanship and archery, and just as practical. As far back as the Heian period (794-1191), swimming was a part of the formal training for the Japanese warrior, and there is no lack of military stories involving swimming skills.

In the Edo Period (1603-1867), the art of Eiho achieved its peak. Several different ryu were already established, and teachers achieved a recognized status.


During the time of the Sengoku, also known as the Warring States period, samurai clans would often battle each other over rivers and seas. In the 15th century until the 17th century, warriors would swim across rivers while in their armor and helmet. In the 17th and 19th century, the samurai would pass messages to each other by navigating through the rivers, seas, and lakes of Japan. With a samurai suit weighing as much as 44 pounds, the warriors were expected to know how to fight on water, or at least to stay afloat.

There are even videos about it: https://www.youtube.com/watch?v=WwDvJeP4WOg&feature=youtu.be

According to wikipedia for richer Samurai the breastplate and helmet where made from metal, so maybe they are comparable to European armors.

TL/DR: It is a skill you can learn.

  • 2
    $\begingroup$ Samurai armor has a lot of laminated wood in its construction, correct? I think that it therefore has a very different effect on the swimming capabilities of its wearer than pure metal armor. $\endgroup$ Jun 17 '19 at 13:59
  • 1
    $\begingroup$ According to Wikipedia: Japanese armour was generally constructed from many small iron (tetsu) and/or leather (nerigawa) scales (kozane) and/or plates (ita-mono), connected to each other by rivets and macramé cords (odoshi) made from leather and/or braided silk, and/or chain armour. German Wikipedia also mentions bamboo plates and that metal armor was save against gunfire in the 17th century. $\endgroup$
    – Fels
    Jun 17 '19 at 14:46

There's a few things to bear in mind for this.

The first is that metal armour is expensive, so it's debatable how many soldiers would actually be wearing it. Clicking around indicates that plate mail would be the equivalent of between one to four years income for a man-at-arms (depending on the period in question - things like the Black Death had a huge impact on labour costs).


And it's heavy both in and out of water - there's a reason knights tended to charge around on horses (above and beyond the fact that they're relatively defenceless when on the ground; there's plenty of tales of dehorsed knights being hamstrung and killed via a knife in the eye by pages and the like)!

Plus, as other people have noted, there's normally a lot of water-absorbent padding under the armour, which would further increase weight and make it harder to perform swimming motions; the extra weight would also stick around after they'd come out of the water.

And while body-fat might help with buoyancy, it's unlikely that many soldiers would be "bulky" enough for it to help. Life was a lot harder, food was scarcer and people were far more likely to have physical defects - for instance, British longbowmen often had skeletal deformities from their training. Plus, you know, army rations are rarely lauded by soldiers :)

So if there's a soldier in metal armour crossing water on foot, it's probably the result of a rout and they've probably been dehorsed and will already be suffering the mental and physical effects of defeat, exhaustion and injury.

Overall, I wouldn't give too much for their chances of making it to the other side...

On the other hand, it's interesting to look at the Roman army, with their standardised armour. A glance indicates that their mail armour (lorica hamata) weighed around 11kg (~24lb). They were pretty successful when it came to marching and conquering, so this may represent the best weight-compromise for marching/fighting/fording/etc.


Even more simple answer - snorkel. Of course it depends on the depth of water, but a few metres of breathing tube would be well within medieval technology. Then it doesn't matter whether you can float at the surface or not.

  • 8
    $\begingroup$ Snorkels are nice, but anything above 1 meter becomes basically impossible to snorkel due to the pressure difference. There is a reason snorkels are usually limited to 30cm. The even simpler answer is a bag of air, or a boat. I have managed to cross small bodies of water with heavy gear by filling waterproof sacks with air (about 30 liters per sack). You will also want a rope, so you can pull it over once you have crossed. Swimming with heavy bags is hard, and would be near impossible with a current. $\endgroup$ Jun 15 '19 at 22:43
  • 3
    $\begingroup$ The question asks how much metal. You did not provide that answer. $\endgroup$ Jun 16 '19 at 5:28
  • $\begingroup$ @DanielVestøl A boat? Why did we not have that idea ... $\endgroup$ Jun 16 '19 at 17:57
  • 6
    $\begingroup$ @HagenvonEitzen I guess that's the ultimate answer. How much armour? About 100,000 tonnes of it, and call it the USS Gerald R Ford... $\endgroup$
    – Graham
    Jun 16 '19 at 21:08

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .