A familiar figure appears repeatedly in textbooks and scholarship alike: that smallpox killed 20–30 percent of those infected. It is a striking statistic—but how reliable is it? Eric Schneider and Romola Davenport have revisited this question.
In 1707, as smallpox spread across Iceland, one observer described a haunting scene: farms stood silent, livestock wandered unattended, and “the healthy could not tend to the sick.” Entire households had fallen ill at once. Some who might have survived, he wrote, died simply because no one was left to care for them. All told some 25% of the population of Iceland died from smallpox in this epidemic. This epidemic shows that smallpox could be very lethal, but was this typical?
In our latest research, we examine smallpox case fatality rates in two eighteenth-century case studies: Iceland in 1707–09 and Sweden in 1776–1800. This allows us to understand how the lethality of smallpox varied in different epidemiological contexts, before the impact of vaccination.
A disease of childhood
In Sweden and many other parts of Europe before vaccination, smallpox was endemic. In such settings, it circulated continuously and was primarily a disease of childhood. By adulthood, most individuals had already been infected and acquired lifelong immunity.
Using detailed mortality data from Sweden between 1776 and 1800, we show that adult deaths from smallpox were extremely rare. This observation creates a puzzle. Given the number of reported smallpox deaths, if smallpox really killed 20–30 percent of those infected, a large proportion of the population should have remained susceptible into adulthood, but they did not.
By modelling mortality and immunity together, we estimate that the most plausible case fatality rate in this endemic context was much lower: around 8–10 percent. A very different picture emerges when smallpox struck as an epidemic disease.
When epidemics overwhelm society

The photo shows how sparse the population of Iceland was even in the capital city Reykjavik in the 1860s. (Sigfús Eymundsson 1837 – 1911, via Wikimedia Commons)
In Iceland, where the population was too small to sustain endemic transmission, outbreaks occurred only intermittently. When they did, they affected both children and adults. The epidemic of 1707–09 was particularly devastating, killing over a quarter of the population. By combining census and mortality data with estimates of infection rates, we calculate that the case fatality rate in this epidemic may have been as high as 43–55 percent.
Such figures are far above the familiar 20–30 percent. But they are not easily explained by biology alone. For historians of medicine, the most revealing aspect of our work lies in its emphasis on care. In endemic settings, infections were spread out over time. Households continued to function, and the sick could be nursed. In Iceland, by contrast, entire households fell ill at once. There were too few healthy people to care for the sick, prepare food or even tend livestock. Under these conditions, mortality rose sharply, not simply because the disease was virulent, but because the social systems that sustained life had broken down.
Beyond a single number
These findings challenge the idea that diseases have fixed case fatality rates that can be applied across time and place. Instead, they highlight the importance of context, especially the organisation of households, the availability of care and the scale of outbreaks.
They also have wider implications. High mortality in past smallpox epidemics, including those in the Americas, may reflect not only immunological vulnerability but also the social disruption caused by widespread infection.
Smallpox, in this light, was not simply a biological phenomenon. Its deadliness depended on the societies it struck. For historians, this serves as a reminder that disease outcomes are shaped as much by social conditions as by pathogens themselves.
Eric Schneider is Professor of Economic History at the LSE. His research focuses on the history of child health and the causes of the health transition. https://www.ericbschneider.com/
Romola Davenport is a research professor in the Cambridge Group for the History of Population and Social Structure, University of Cambridge. Her research addresses the demographic impacts of early public health interventions. https://www.geog.cam.ac.uk/people/davenport/
Their open-access paper is available at Schneider, E. B., & Davenport, R. J. (2026). What is the case fatality rate of smallpox? Population Studies, (ahead-of-print), 1–15. https://doi.org/10.1080/00324728.2026.2620692







