How long was the average Roman foot, and what was their shoe size?

Archaeologist Eric Poehler just keeps coming with the questions about Roman walking and feet.  Today, he wanted to know the size of the Roman foot.  In my last post, I'd kind of given up on the idea of figuring out foot size, since I didn't think I had any foot measurements.  Then I remembered this morning that of course I have calcaneus maximum length.  The trick was to find a formula using calcaneus maximum length to approximate foot size.

Sandaled foot from the Augustan period (Met Museum)
This was more difficult than you'd think.  There are a metric TON of articles that relate shoe/foot size to stature, but you have to have the shoe/foot (these are useful in forensic contexts, of course).  So I could use long bones to calculate stature and then use stature to calculate approximate foot size, but that would introduce one more level of error than I need.  It seems like no bioarchaeologists care about estimating foot size from foot bones, which surprised me because I'd assumed at least comparative primate morphologists would be interested in this.  (Now, of course, there is growing interest in Roman walking because of databases like Stanford's ORBIS.)

I did find what I was looking for, though, in literature related to lower leg changes in polio: Anderson, M., M. Blais, and W.T. Green. 1956. Growth of the normal foot during childhood and adolescence. Length of the foot and interrelations of foot, stature, and lower extremity as seen in serial records of children between 1-18 years of age. American Journal of Physical Anthropology 14(2):287-308.  This article is helpfully available for free via PDF here.  If you scroll through to page 306, there's a handy chart that gives you the percentage of the foot made up by the "calcified os calcis" (what we now simply call the calcaneus).  Taking 18-year-olds (as the authors have concluded that the foot is no longer growing at this point), we find that the calcaneus makes up 30.2% (+\- .1) of the male foot and 28.9% (+\- .1) of the female foot.  Spiffy!

Now, we take the maximum calcaneus lengths from the population I studied at the Imperial-era Casal Bertone cemetery, 2km east of Rome. The male average was 80mm (8 cm), and the female average was 75mm (7.5 cm).  Using the power of multiplication, the average Roman male foot was 26.5cm, and the average Roman female foot was 25.9cm.  If you want to go a step further (ha!), this means the average Roman male from Casal Bertone wore a US 8.5 / EU 42 shoe, and the average female a US 10 / EU 41 shoe.  Boom -- calculating calcanei!

The female numbers seem too long, honestly, but I can believe the male numbers.  If you recall my previous post, the average male stature from this site was about 167cm, and female stature was 157cm.  So (using Imperial measurements now, sorry, but I'm American!) a 5'6" man could easily wear an 8.5 US shoe.  But a 5'2" woman would not wear a 10 US shoe.  I'm 5'9" and I wear a 10.

If there really aren't equations other than this to approximate foot size from calcaneal length, I suddenly have an MA project in mind for an interested student... And the correlations between bone length and shoes (as from Vindolanda) have lots of potential as well!


magistra said…
I am a 5"2' woman and I wear a size 5. I have a terrible time finding shoes in the US, but Italian (read Roman) shoes fit beautifully. I am convinced that a Roman woman would wear between a 5 and 6, maybe a 7, but no larger.
And yet I've heard from a couple women who are much shorter than I am and wear the same shoe size. The correlation between stature and shoe size isn't perfect, clearly. But I think we need some better formulae for predicting foot size from bones.
Unknown said…
"So (using Imperial measurements now, sorry, but I'm American!) a 5'6" man could easily wear an 8.5 US shoe. But a 5'2" woman would not wear a 10 US shoe. I'm 5'9" and I wear a 10."

I am not sure I understand this comment because in the U.S. a man's size 10 is not a women's size 10 (in U.S. and Canada the size number for a female is 1.5 larger). So, I think your numbers work because a woman's size 10 is a man's size 8.5. In your post you say that in European sizes the male would have worn a size 42 while a female size 41. This is great, because in the size conversion, the male would be U.S. 8.5 while the female would be U.S. 9.5, just a half size smaller. See for the chart I used.
Yes, of course men's and women's size 10s are different.

Let's take Americans again as an example. American women are, on average 5'4" and wear a size 8 shoe. American men are on average 5'9" and wear a size 10.5. (Quick info gotten from here:

A Roman male was about 5'6" on average, so wearing an 8.5 men's shoe seems perhaps small but pretty right.

A Roman female was about 5'2", shorter than the average American woman. But the foot size calculation has her wearing 2 sizes larger. That doesn't make sense. Even a 9.5 US women's doesn't make sense.

I will agree, though, that the European shoe size system is easier than the confusing American one. (The Japanese system of just using flat-out centimeters is even better, though.)
Unknown said…
It might be unusual, but I am a woman tall 159cm and my European shoe size is 41, so I would believe it possible.

Interesting article, thank you.
Unknown said…
I am a woman 159 cm tall and my European shoe size is 41, so I would think it quite believable.

Interesting article, thank you.
PetridisKrohn said…
The shoes found in Vindolanda er in size 33 for women (at a maximum), so this sounds a bit off. Either that or the intepreration of women in Vindolanda needs new evidence. The idea is good though. I'm doing a paper in roman footwear from Vindolanda, where I'm trying to determent the identity in military camps with the footwear as my archaeological materiale.
Unknown said…
Romans developed feet and inches miles etc
It is implied that a foot in measurement is based on the size of the average soldiers foot, that would make the average solider size 15.5 UK is this right??

Popular Posts