Calculation of the European population
in the 5th century
The basic rule is that the population in the 5th century AD was some 10% of the modern one. In fact the European
population hardly changed over the ages until the 19th century. This figure must be adjusted according to the fertility
of the land and the climate. For southern Europe we must increase the figures, for northern Europe we must decrease the
France has today some 65 million inhabitants. In AD 428 the population would be 6.5 million. But as Gaul or France always
was a fertile land with a heavenly climate, we must increase this figure, 8 million is more realistic. Denmark has today
some 5.5 million inhabitants, 10% = 550 000, but as the climate is cold and wet some 300 000 people is closer to reality,
We must take in account a number of criteria:
(1) The staple crops were of Mediterranean origin. This means that the wheat varieties were not yet acclimatized
(or selected for) to northern climate conditions. Even today, this acclimatization is not complete. At the time, a bad
summer in northern Europe meant a considerable drop in production, if not famine.
(2) Farmers used a system of production spread over 3 years on the same land. The first year the most valuable but
also the most demanding crops were planted, like wheat. The second year, less demanding crops were planted, like some
vegetables. The third year, the land rested. At the end of the third season, what ever grew on the parcel was burnt. The
ashes fertilized the land and the next year exploitation could start all over again. This system meant that 3 times more
land was needed to feed the same population. As not all the land was as fertile, we can estimate that 1 family (6 to 10
persons) needed some 30 hectares arable land to feed itself. 10 hectares laid at any given
year in waste. 10 hectares gave in theory sufficient crops, given the fact that today, with the help of modern
fertilizers and seed varieties, the yield is at least 10 times greater. So, 1 hectare produces today as much as 10
hectares in the 5th century.
(3) Meadows were as much needed. Cattle was very important. It provided a secondary source of food and income,
especially when the staple crops failed. Horses were the only mean of transport and the most important source of power.
10 hectares of meadows per family is a conservative estimation. Most of time, it was land
that was less suited for growing crops.
(4) Woodland was also very important. Wood was the main building material and was the only mean to warm the farm.
Here too, at least 10 hectares of forest seems necessary, although it stood upon the least
valuable grounds. It's a common mistake to believe that there were more forests in Britain 2000 years ago than today. The
reality was that the farmers had chopped all woods upon fertile and arable land. One cannot eat wood.
All land was exploited up to its maximal sustainability. The idea that vast virgin forest subsisted in western Europe is
false. 4000 years of agriculture managed to transform the landscape profoundly .
In total, 1 family needed an estimated 50 hectares to subsist. The 50 hectares, a mix of
arable land, meadows and woods, is a theoretical average. This means that in some regions the woods would dominate, while
in other regions arable land would. Fertile soil meant that less land was needed, let's say some 20 up to 30 hectares.
The local climate is also very important.
The United Kingdom has today a surface of 250 000 km2. In hectares: 25 million. Divided by 50 (per family)
gives: 500 000 families. In number of persons: between 3 million up to 5 million. The truth is in between: some 4 million
is a good estimation. Not all of Britain is suited for farming. The Scottish highlands are much less suited than the
southeast of England, or the Salisbury plains. One can estimate that only 1/3 of Britain is really suited for agriculture
(arable land and good meadows).
This leads us to the following rule : The European population 2000 years ago was in general some 10% of the modern
one. This figure must be corrected per region. In the colder North one can subtract around 20% from the amount, in the
warm South one can add 20 %. In Scotland and Scandinavia the actual population figures were considerably lower.
The calculation for Italy = 60 million today = 6 million at the time of Julius Caesar, but given the excellent
climate which is well suited for Mediterranean crops, we can add 20% = 7 up to 8 million. Italy always was the
most fertile and richest region in the Mediterranean basin. The real power of Rome was derived from superior agricultural
output in Italy.
How can we verify those figures? By checking with the size of the professional armies the Romans disposed of. A
militia represents between 6% up to 10% of the total population (this population includes of course all women and
children - 35% was younger than 20 years). 6% is the figure for the first battle, 10% is reached when the war lasts for a
long time (repeated recruitments). A professional army represents 10% of the possible militia (or 0,6% - 1% of the
population). Those figures are valid for the whole world and any period. Let's not forget that the occupation of a
soldier is very demanding, especially in ancient times .
As Italy had an estimated population of 7 million, so the possible professional army, at a rate of 1 % is 70 000 men. Or
in legions: 70 000 / 6 000 = 11,67 or 12 legions. But the Romans recruited also outside Italy. Rome's army was in fact
bigger (estimated at 120.000 men). The real recruitment percentage was smaller (0,6%) and more recruitment happened
In 48 BC, after the war in Gaul, Julius Caesar had an army of 8 legions or 8*6000= 48 000 men (some speak about 11
legions), although many legions had not been repleted. Several legions like Legio V Alaudae also called
Gallica (Southern France) and Legio III Gallica , were entirely recruited amongst populations outside
Italy. Many others had a mixed recruitment base, even with south-German (or Austrian) volunteers in the cavalry
Their training and war experience had made them very skilled. No wonder Caesar had ambitions.
The same calculations can be applied to the Belgian Nervian army, a militia. The tribe controlled some 10 000
km2 or 1 000 000 hectares (1/3 of modern Belgium). As the country is flat and fertile, we take 30 hectares per
family. Total estimated population: 260 000 people. The militia at a rate of 7 % = 18 200 men. A professional soldier is
generally worth 2 militia men. The reason is his superior weaponry, experience and discipline. Most Gauls fought with
hunting weapons (said Caesar). Only 1 in 10 had a full war gear. Caesar’s army counted at the moment of the
battle of the Sabis (against the Nervians), some 44 000 fully equipped professionals. To be a match for this army,
the Nervians needed 70 000 men, even taken in account their courage. A numerical superiority of 30% is a minimal
requirement, or some 59 000 men. Caesar wrote that a part of the Nervian army had to remain in their homeland to counter
the German raiders he had sent. So, an optimistic figure for the Nervian army is that they had some 13 000 men on the
banks of the small Sabis river. 2 other tribes assisted them with in total an approximate 6000 men. It becomes clear that
the Nervians had little chance to win, even if their army had been entirely professional. Nevertheless, Caesar almost
lost that battle. The Romans counted many casualties that day.
The figures of the Belgian armies which Caesar gave us, are clearly exaggerated. According to these numbers, all the
(mentioned) Belgian armies counted some 268 000 men in total. Were those figures to be correct, then Belgium had
according to Caesar a population (I take 6%) of 4.5 million, that is, more than Britain. Twice as much as the Belgian
army in 1914.
How big was Rome at its peak?
How many inhabitants lived in Rome itself during the first century AD, the age of greatest power? Rome was without a
doubt the biggest city of Europe. But how big?
Recently the accepted population figure has been one million. That number is simply impossible, even preposterous. Think
about it: how can you bring in food for a city of one million inhabitants with the technology the Romans had? Paris and
London did not reach their millionth inhabitant until shortly after the development of the railways. Trains can transport
food fast, fresh and in great quantities. They boosted the growth of European cities. Rome had no trains, so imagine the
trail of carts. The staple food was bread. We know that wheat was transported by ship, often from Egypt. Wheat is one of
the easiest things to transport. However, wheat had to be ground into flower. In ancient Pompeii such a thing was done
with donkeys which turned around the mill stone for hours on end. We guess that Rome was no different. One million people
ate some 500g average per day per person because work was very physical and anyway, other food was far more expensive.
This gives us about 500 tonnes per day. All this flour had to be baked too. The problem here is the quantity of wood -
charcoal actually - necessary to heat the ovens.
But people liked meat very much, although the poor rarely ate it. I propose that 200000 people regularly ate meat, some
200 grams per day. I calculate here some 40 tonnes per day. Meat entered the city on foot. Without fridges, meat rots
very quickly. Let us take goats or sheep as 'staple meat'. Sheep or goats are not choosy when it comes to food: all
greeneries are good. Pigs, however, need more energy-dense food, so were more rare. Cows were the exception. The figure
in the British Domesday book (AD 1086) gives one cow per 40 sheep/goats. One goat gives about 25 kilo meat, thus some
1600 goats or sheep per day had to be brought in the city and slaughtered. Can you imagine the pile of rotting guts after
a week? We speak about 280 tonnes of stinking material. Is that possible? Let us take a circle of 150 km around Rome.
Such a region would correspond roughly with modern Holland. Holland has today some 350000 goats. If the region around
Rome was as big as Holland and as productive then the supply of goats would have lasted only 220 days. Not even one year
and all goats in a circle of 150 km were eaten. I doubt that Latium, the region around Rome was as fertile as Holland is
today. Latium is much smaller too: a circle of some 40 km around Rome.
So, one can argue that the goats could have been brought in from further away. Possibly, but the rate of reproduction of
a goat is not that spectacular. To fill the gap, a region needed to be at least three times the size of Holland. In
reality, given the much less rate of agricultural productivity, one has to multiply this 'region' by seven, which means
that roughly the whole of Italy (excluding 75% of the mountains) had to breed sheep or goats, just to pleasure the
Romans. The Italians themselves, although Romans citizens too, would be forbidden to eat sheep/goat meat. So you see the
problem: food production, whatever it is, was simply inadequate to feed 'one million citizens' Rome.
We know nevertheless that the Gauls managed to send geese to Rome. In a similar way, Britain was able to send oysters to
Rome. But these were exceptions and that is the very reason why we know: nobody writes about the daily routine.
We desperately tried to find out who is responsible for the one million figure. Historians refer as usual to each other,
so the circle seems to be round. We could not find the original source. Anyhow, we challenge anybody to prove that
feeding one million inhabitants with the technology of 2000 years ago is possible.
In the mean time, we would like especially like American historians to be more cautious when pretending that ancient Rome
had one million inhabitants. Given its surface, 13.7 square km intra muros (->Wiki), it supposedly meant at
least 73000 people per square kilometre, a figure even modern Shanghai with its forest of sky scrapers does not have, let
But then how to calculate its real number of inhabitants?
First method: In 1800 the average percentage of city-dwellers in Western Europe was 10% of the total
population. This figure is well attested. During the Roman Empire I can estimate this percentage at 5% average. The land
had to feed the cities, so the more food the land was able to produce, the more city-dwellers were possible. The yield of
the land in AD 50 was probably half of the yield in 1800 as the total European population at that moment also was double
the figure of AD 50. Example: France under Napoleon had 20 million inhabitants. The reason is a much better agricultural
technology in 1800. Therefore: 5% in AD 50.
Italy had during the first century an estimated population of 7 million. The country was the richest and therefore the
most populated one of the Mediterranean region. Italy was the power engine of the Empire. The estimated total city
population at 5% = 350 000. Pompeii had an estimated population of 10 000. There were at least 12 cities with this number
of inhabitants in Italy (=120 000). Some cities like Mediolano (Milan) had probably more inhabitants :± 20 000
. So: 150 000 citizens in the important cities. 350 000 - 150 000 = 200 000 left for Rome. But there were also
a lot of smaller cities with something like 5000 inhabitants. In fact, Italy has countless villages. Let's say: 30 = 150
000 total lived in the villages. 200 000 - 150 000 = 50 000 left for Rome (estimated error margin: 20%). But we know that
Italy was not able to feed Rome. Grain had to come from Egypt for instance. It was mainly the Mediterranean region which
fed Rome during the first century AD.
At its greatest extend the Roman Empire counted some 45 - 50 million inhabitants. North of the Pyrenees lived an
estimated 16 million people (Gaul: 8 million +Britain: 4 million +, the Germano-Romans (from Belgium to Austria): 4
million = 16). Before the Roman conquest of those regions, the Empire had some 45-16 = 20 up to 23 million inhabitants.
We must subtract Italy itself: 20-7 = 13 or 14 million. 14 million at 5% = 700 000 city inhabitants in the regions
If we add 100 000 inhabitants to Rome (50 000 + 100 000), then those 100 000 represented 14% of the total of all cities
under the rule of the Roman republic except Italy (100 000 people of 700 000 = 13%).
If we add 200 000 then it's 28%, or almost a third of the total city population around the Mediterranean Sea, except
Italy. Then the yield surplus must have been greater than 5%, at least some 7,5% - which is unlikely given the rather dry
climate. Only fertile Egypt was capable of such a thing, but not every single year. As Egypt had some 5 million
inhabitants, (3000 000 hectare at some 1,7 persons per hectare), the calculation is: 7,5% of 5 million= 375 000. To able
to feed the 100 000 citizens extra in Rome meant that almost one third of the Egyptian surplus was destined for Rome.
When Marcus Antonius halted the shipping of grain to Rome, the food situation in Rome rapidly deteriorated.
In theory, other regions could have supplied Rome too, but none were as fertile as Egypt. North Africa could at best feed
some 40 000 extra Roman inhabitants. What becomes very clear is that Rome never could have more than 250 000 inhabitants.
Given the fact that harvests were not always optimal in Egypt and North Africa, I can add 120 000 inhabitants to the
initial 50 000 .
Conclusion of the first calculation method:
Total estimated population of Rome during the 1st century was about 170 000. The error margin is a sketchy 20%.
A second method to estimate the number of inhabitants is to calculate the surface of Rome and to compare it
with a medieval city. This medieval city is the rich city of Bruges. The assumption is that building technology
was similar, as was the density of the population. Medieval Bruges had an attested 40 000 - 45 000 inhabitants. Its inner
surface is some 430 hectares. This gives us an average of 45000/430 = 105 inhabitants per hectare ( one hectare is
about 2.5 acres). The Aurelian city walls of Rome were build in AD 271. The total area within the walls is some
1370 hectares. Rome had thus an area 1370/430= 3,18 times bigger than Bruges. Its population can therefore be
estimated at 45 000 * 3,18 = 145 000. Again, an error margin of 20% has to be considered. Given the fact that Rome
had, in contrast to Bruges, apartment buildings, we can estimate that this figure corresponds with the previous
A third method
is to compare Rome with Pompeii
. Pompeii was 66 ha big and had 10,000 inhabitants
(Wikipedia), there seems to be some sort of consensus there.
Actually, Pompeii looks like a egg where the southern half was shaved off.
To the west, one can still see the ancient beach, but I noticed that the amphitheatre is situated almost upon the southern
city border, which is highly unusual. Such an infrastructure was usually situated more in the centre. So, Pompeii could
have been bigger than the 66 visible hectares.
Anyhow, 10,000 / 66 = 152 inhabitants per ha. If we missed the southern part, perhaps an extra 40 ha , then 10,000 / 106 =
The likely assumption is that the covered part of the city is similar in density to the uncovered part. I wonder how
archaeologists came to the figure of 10,000 for only half of Pompeii was dug up.
Anyhow, let us make a compromise: "Pompeii had 125 inhabitants per ha."
Rome had 1370 ha intra muros. Smaller cities had typically a greater density, but Rome had apartment blocks. So we can
assume a similar density in Rome. That gives us: 125 x 1370 = 171,250 inhabitants. Given some 20% margin more of less.
This insula at Ostia had only two stores.
Fourth method: counting the number of houses. Rome had 1370 hectares, 44300 insulae (apartments) and 1781 'houses of the rich' intra muros.
I can not take in account what was supposed to be outside the wall. Walls around a city are in general build slightly too
big. But that is a different matter. A city need streets. I guess that 1/3 of the space, at least, is needed for streets
and squares. 2/3 of 1370 ha or 13 700 000 m2 = 9 133 333 m2 available for insulae. A typical insula
measured some 300-330 m2 (ref: Wiki) so 9 133 000/330 = 27 676 insulae possible. Wikipedia gives a 'typical' 6
to 7 apartments per insula. I doubt for such a number is unproven. Rome had hard soil, rock underground but also soft,
muddy soil. 6 to 7 apartments is good when the insula is build upon hard soil.
Accepting this means : 6 * 27 676 *2.4 = 398 534 inhabitants. Still not a million. But I saw a map of Rome and it seemed
to me that only half of it was used for popular living quarters. Some private gardens intra muros were huge. So: 1370/2 =
685 ha or 6 850 000 m2. 6 850 000/330 = 20 758 insulae. I take 4 apartments as standard average: 83 030
apartments x 2.4 = 332 121 inhabitants. My conclusion here is that a figure of 44300 is hard to believe in any case.
However, according to Wikipedia (insula) the word insula could mean apartment or apartment block. This changes the
calculations completely. If insula meant single apartment then 44500 * 2.4 = 106.800 inhabitants.
I previously put forward some 160 000! 160 000 / 44 500 = 3.6 or 4 persons average per apartment which sounds very
reasonable. All these calculation suppose of course that all the insulae were occupied. However, this is far from
certain. On top of that, it is possible that the 1781 'individual houses' or 'richmen's houses + gardens' took more space
than I supposed until now. Anyhow, I accept to go up to 250 000 inhabitants. 250000 / 44500 = 5.6 or 6 persons average per
apartment. But, not all apartments were occupied. In compensation, we can add the 'rich people' and
their slaves to the number of insula-inhabitants. Although it is possible that the slaves simply lived in the insula next door. It remains guessing.
Note that (about) 6 x 6 m was the standard room size and at the same time a house width and depth in medieval Bruges.
This must have something to do with the length and thickness of the standard core-oaken 'mother' beams (a foot x a foot
thick) which are strong enough to support 'sufficient' weight. I think that such beams were what one could get from an
average 'ripe' oak. In Bruges two rooms were needed: one to live/work in on the ground floor and one to sleep in on the
first floor. An attic of the same size as storage room was also present: 36m2+36+36= 108 m2 total
living space. It cannot be a coincidence that this corresponds with a Roman apartment of 75-100 m2.
First method :
Second (Bruges) :
Third method (Pompeii):
Fourth method (insulae):
It is clear that Rome could not have one million inhabitants.
At the time of Julius Caesar, 50 BC, Rome was had say 80 000 - 90 000 inhabitants. The area within the republican walls
(until 40 BC) is also known and at least half the size of the city in AD 271.
Consider the historic statistical average that the number of professional soldiers represent something between 0.5% up to
1.5% of the total population. For belligerent Rome itself I'm prepared to go up to 5 % (almost the size of a militia).
This means that the city of Rome itself was capable of delivering only one single legion (=5000 legionnaires) at the end
of the republic! As the total Roman army during the age of Caesar counted some 120 000 soldiers, then we can only
conclude that the Romans recruited the bulk of their soldiers outside the city. At least 90% of the Roman soldiers were
therefore NOT from the city itself but were Italians, South-Gauls, Spaniards, South-Germans, Greeks, etc.
It is also a common mistake to believe that the Roman senators originated from the city itself. 98% of the senators came
from other regions, but the idea was that they could co-rule the city. Ruling the city meant ruling the Empire. This
was the Roman system: the incorporation of foreigners into the center of power. This brought stability and loyalty.
That's one very important reason why the Empire lasted so long. The Roman army was based upon the same system, 98% of the
legionnaires were recruited outside Roma city.
Check: When the Empire fell, Rome had an estimated population of 50 000 - 60 000. The city needed no longer to import
grain from Egypt.
How big could Londinium have been?
In theory, Britain could feed some 150 000 city inhabitants. Considering the not always ideal climate, it must have been
closer to 100 000. This represents 50 villages with 2000 inhabitants. Some cities were bigger. London, Caerleon,
and a few others had closer to 4000 inhabitants.
The wall around London was build for growth within. All similar walled cities were build with that in mind, including
medieval Bruges. The orange zone above represent the probable rows of houses.
The total surface intra muros was about 124 ha (my calculation). Which gives us 18,600 max. inhabitants if the
density of Rome is supposed.
This figure is unacceptable for it supposes a much higher density of streets, similar to the modern pattern. The modern
part of the city, roughly between St Paul's cathedral and the London Tower - which of course are of a post-Roman date -
has clearly far more streets. I haven't counted them but a tenfold is certainly the case. We must keep in mind that space
was still in large supply at the time.
What I see is a paucity of streets.
What I suspect is that people build their houses alongside the roads. A lot of space remained unused. This extensive use
of space in London happened before the walls were build. After all, London was 'lands' and a 'land' was a house with its
Given the streets on the map, I calculated some 620 houses, taking for each a width of about 6 metres. The calculation
gave about 4000 inhabitants as a result, take or give a few hundreds. Well, I am prepared to concede a few thousands
inhabitants more. But that was about it.
Maybe 5 or 6 cities had more than 2500 inhabitants in Britain. And what about the many villages? In reality Britain had
more 'cities', which makes the average size of a British town smaller. The good news is that Britain had indeed a fair
number of villages. The bad news is that Britain's cities were too sparsely populated and as many city dwellers went back
to the countryside after 250 AD. The British cities became gradually less populated from that time on. When the
Anglo-Saxons arrived, they found most cities abandoned.
Compare : in 1821 the population of Sheffield had reached 31 314 inhabitants. Today it's close to 530 000.