How many died? It sounds like the simplest question about the Black Death, and it is one of the hardest. No medieval state counted its people the way a modern census does. Historians have to estimate deaths from records kept for other purposes, and the answers have changed as new evidence has come in.
Why the chroniclers mislead
Medieval writers gave plenty of figures, and most are useless as counts. A chronicler might say that nine out of ten died, or that a city lost a hundred thousand people when it never held that many. Chroniclers often used round, inflated figures to express horror, not to measure it. Numbers like a thousand or ten thousand meant "very many." The writers were also in the middle of the disaster. The Florentine chronicler Giovanni Villani died in 1348, and Villani's chronicle broke off when he died of plague, leaving a blank where he meant to record when the epidemic ended. His brother Matteo took up the work.
Records never meant to count deaths
Better evidence comes from documents kept by people who needed accurate lists. English manors held regular courts that recorded land changing hands. When a tenant died, the lord often took a heriot. A heriot was a death duty paid to the lord, often the dead tenant's best animal. Each heriot marks a death, and each transfer of land to an heir marks a household that went on. By following names through these rolls, the historian Zvi Razi estimated that roughly 40 percent of the adult men on the manor of Halesowen, near Birmingham, died in 1349.
The church kept records too. When a parish priest died, the bishop had to appoint a successor and write it in his register. Bishops' registers are useful because they record new priests replacing dead ones. In several English dioceses, these registers suggest that around 40 percent of the parish clergy died. Priests visited the sick and dying, so they may have died at higher rates than their neighbors, but they were also better fed than most peasants.
How high was the toll?
For a long time the standard estimate held that about a third of Europe died. More recent work has pushed the figure up. Ole Benedictow estimated that about 60 percent of Europe's people died, though many historians think his figure too high for some regions. Local evidence varies enormously. Some Italian cities may have lost half or more of their inhabitants in a single season.
New kinds of evidence show that variation. Pollen trapped in lake beds and bogs records which plants grew nearby, so a fall in cereal pollen and a rise in weeds and trees can reveal fields being abandoned. A pollen study published in 2022 found that death rates varied greatly between regions. Farming collapsed in much of Italy, France and Greece, while in parts of Poland, Lithuania and Ireland it carried on or even grew. The Black Death was a continent-wide disaster, but not an even one.
Beyond Europe
The Middle East and North Africa suffered as badly. Egyptian and Syrian writers describe streets of empty houses and bodies carried out by the hundreds each day. Historians estimate that Egypt and Syria lost perhaps a third of their people in the first wave, and repeated outbreaks over the following century kept the losses high. For India and China the evidence is too thin to say how hard, or even whether, the Black Death struck.
Waves that kept returning
The first wave was only the start. Plague returned in 1361 to 1362, and English chroniclers called it the mortality of children. The outbreak of 1361 struck children and the young hardest, probably because they had been born since the first wave and had never been exposed. More outbreaks followed in 1369, 1375 and on through the next century. Europe's population stayed low for over a century because plague kept returning every decade or so. England, which may have held five or six million people before 1348, held perhaps two and a half million by 1377, when a new poll tax listed everyone over fourteen.
Worked example: reading a court roll
Imagine a manor court roll from the spring of 1349. In an ordinary year the court records perhaps two or three deaths. This session lists twenty-two, each with a heriot: an ox from one holding, a cow from another, and, from several, nothing, because the tenant left no animal. Eight holdings pass to sons or daughters, nine to brothers or nephews, and five are "in the lord's hand" because no heir came forward.
What can the historian conclude? First, a death rate: if the manor had around fifty tenant households, twenty-two deaths among the heads of household means over 40 percent. Second, the cases with no heir, and the long chains of inheritance through distant relatives, show families wiped out or scattered. Third, the roll counts only tenants, mostly adult men and widows, and says nothing directly about children, servants or the landless poor. A careful estimate states that limit openly instead of pretending the roll covers everyone.
