Ebola: The Deadliest Outbreaks and What Stopped Them

Ebola virus disease kills faster than almost any pathogen that has caused a major outbreak in the modern era. Death rates have ranged from 25% to 90% depending on the species and setting, according to the WHO Ebola fact sheet. Every Ebola outbreak before 2026 was eventually stopped. The Bundibugyo epidemic in DR Congo, the second-largest on record, is still running.

Understanding how outbreaks end requires looking at the full arc, from the virus's discovery in 1976 to the ring vaccination strategy that changed the response in the late 2010s, and the 2026 epidemic that exposed its limits.

How was Ebola discovered?

In September 1976, a mysterious hemorrhagic illness appeared near the Ebola River in what was then Zaire (now the Democratic Republic of the Congo). The village of Yambuku and its surrounding communities saw 318 cases. 280 people died, an 88% case fatality rate, according to CDC's outbreak history.

Belgian microbiologist Peter Piot was 27 years old when blood samples from a Belgian nun who had fallen ill in Zaire arrived at the Institute of Tropical Medicine in Antwerp. He and his colleagues found a virus that looked like Marburg under the electron microscope but was clearly something different. It was named after the Ebola River rather than Yambuku itself.

A simultaneous outbreak in Nzara, Sudan, was caused by a different species, now called Sudan virus. CDC records 284 cases and 151 deaths there. Between the two outbreaks, the world had its introduction to Ebola: 602 cases, 431 deaths.

Why were outbreaks small for 40 years?

After 1976, Ebola appeared sporadically in central and east Africa but never spread widely. Before 2014, no outbreak had passed 425 cases. That record belonged to Uganda's 2000 outbreak of Sudan virus, which killed 224 people. Several outbreaks were contained at under 100 cases.

Geography played a major role. Ebola's animal reservoir, likely fruit bats of the Pteropodidae family, lives in remote tropical forest regions. Spillover events typically happened in small, isolated villages. Patients often became too ill to travel far. Burial practices that involved washing the dead, a major transmission route, kept the virus circulating within tight community networks.

Isolation and contact tracing, the basic tools of outbreak response, worked because the outbreaks burned through small communities and ran out of susceptible contacts before reaching cities. The virus's brutality was, paradoxically, a containment advantage.

What went wrong in West Africa?

In December 2013, a toddler died in the village of Meliandou in southeastern Guinea. Investigators later linked him to the outbreak as its presumed first fatality, as described in the first published account in the New England Journal of Medicine. By the time Ebola was identified as the cause in March 2014, it had already spread to Guinea's capital, Conakry.

What followed was catastrophic. CDC counts 28,610 cases and 11,308 deaths in Guinea, Sierra Leone and Liberia. The numbers dwarfed every previous Ebola outbreak combined.

Three factors turned a village outbreak into a regional disaster.

Broken health systems. All three countries had very few doctors for their populations, and clinics lacked basic supplies. Health workers had no training in hemorrhagic fever protocols. Patients showing up at understaffed clinics infected healthcare workers, who then infected other patients.

Distrust of authorities. Decades of civil war in Sierra Leone and Liberia had shattered trust between communities and government institutions. When health teams arrived to isolate patients and safely bury the dead, many communities resisted. Some believed the teams were actually spreading the disease. Violence against response workers occurred repeatedly.

Delayed international response. MSF (Doctors Without Borders) raised the alarm in March 2014. WHO didn't declare a Public Health Emergency of International Concern until 8 August 2014, five months later. By then the outbreak was growing exponentially. An independent Ebola Interim Assessment Panel later concluded the declaration had been significantly delayed, a finding WHO itself did not fully accept.

Containment ultimately required a military-scale response, with treatment centers built with help from US, UK and French forces, and massive safe burial campaigns.

How did ring vaccination change the game?

During the DRC's 10th Ebola outbreak in 2018-2020, responders deployed a new weapon: the rVSV-ZEBOV vaccine (brand name Ervebo), using a strategy called ring vaccination.

Ring vaccination doesn't attempt to vaccinate an entire population. Instead, when a new case is confirmed, response teams identify every person who had contact with the patient, and then every person who had contact with those contacts. This "ring" of people receives immediate vaccination. The strategy creates a firewall of immune individuals around each case, cutting transmission chains without needing to reach everyone.

The concept dates back to the smallpox eradication campaign of the 1960s and 1970s. Applying it to Ebola required a vaccine that worked fast. In the Guinea ring vaccination trial, published in The Lancet, no one vaccinated immediately developed Ebola 10 or more days later, against 16 cases in the delayed group. That gave an estimated efficacy of 100%, with a wide confidence interval of 68.9% to 100%.

During the 2018-2020 DRC outbreak, which affected conflict zones in North Kivu and Ituri provinces, more than 303,000 people were vaccinated. Despite extraordinarily difficult conditions, with armed groups attacking treatment centers and widespread community resistance, the outbreak ended at 3,470 cases and 2,287 deaths, WHO reported when it declared the outbreak over on 25 June 2020.

What tools exist now?

Ebola response has improved dramatically since 2014, but only for one species. WHO lists two approved vaccines: Ervebo (rVSV-ZEBOV) by Merck, a single dose recommended for outbreak response, and the Zabdeno and Mvabea regimen by Johnson & Johnson's Janssen. Both protect against the Zaire species, which has caused the most lethal outbreaks.

Treatments have also advanced. In the PALM trial in DRC, 28-day mortality was 33.5% with the antibody cocktail REGN-EB3, now Inmazeb, against 51.3% with the comparator ZMapp. The single antibody MAb114, now Ebanga, cut mortality to 35.1% against 49.7%. The FDA approved both in 2020, again for Zaire ebolavirus only.

Still, challenges remain. DR Congo has now recorded 17 Ebola outbreaks, including one in Kasai province in 2025, according to ECDC's week 42 threats report. Maintaining cold-chain storage for vaccines in remote equatorial forest settings is logistically brutal. Contact tracing requires community cooperation, which is hardest to achieve in exactly the areas where outbreaks tend to occur: regions marked by poverty, conflict, and institutional distrust.

What is happening with Ebola in 2026?

In May 2026, DR Congo and Uganda reported an outbreak caused by Bundibugyo virus, one of the three Ebola species known to cause large outbreaks and one with no licensed vaccine or treatment. WHO declared a PHEIC on 17 May. As of 20 October, DRC had 9,704 confirmed cases and 4,687 deaths across eight provinces, per WHO's 23 October Disease Outbreak News.

Uganda contained its 20 cases, and WHO and Africa CDC declared its outbreak over on 27 August. Kenya has nine cases in one cluster traced to an imported patient, and Nigeria confirmed one imported case on 20 October.

This epidemic has turned the lessons above inside out. Ervebo was built for the Zaire species, so the ring vaccination that helped end the 2018-2020 outbreak is not a proven option. A vaccine trial began in Bunia on 23 October 2026, and antibody and antiviral treatments are being tested in a WHO-sponsored trial. Until they report, responders are back to the older tools: isolation, contact tracing and safe burials.

How does PandemicAlarm rate Ebola?

PandemicAlarm assigns Ebola virus disease a baseline severity rating of 4/5. The high CFR, lack of widespread population immunity, and limited treatment access in endemic regions all contribute to that rating. The 2026 Bundibugyo epidemic also sits at Level 4 on the live map, reflecting cross-border cases and a declared PHEIC. Our guide to disease severity scoring explains what would move an event to Level 5.

Ebola is not likely to cause a global pandemic on the scale of influenza or COVID-19. It spreads through direct contact with bodily fluids, not through the air, which limits its transmission efficiency. But every new outbreak tests local and international response capacity. West Africa proved that a delayed response to even a "familiar" pathogen can cost thousands of lives, and 2026 shows what happens when the familiar tools don't fit.

Track active Ebola events on the PandemicAlarm map and read more about how related hemorrhagic fever viruses behave in our Marburg virus explainer.

FAQ

What was the deadliest Ebola outbreak?

The 2014-2016 West Africa epidemic, caused by the Zaire species, remains the deadliest. CDC counts 28,610 cases and 11,308 deaths in Guinea, Sierra Leone and Liberia. The 2026 Bundibugyo epidemic in DR Congo is the second-largest on record, with more than 4,600 deaths by October 2026.

How many Ebola outbreaks have there been?

There is no single official count, because small spillovers are not always detected or reported. DR Congo alone has recorded 17 Ebola outbreaks since the virus was discovered there in 1976. Uganda, Sudan, Gabon, Guinea, Sierra Leone and Liberia have also had outbreaks.

How is an Ebola outbreak declared over?

Health authorities wait 42 days with no new confirmed case, counted from when the last patient leaves care. That is twice the maximum 21-day incubation period. If the window passes cleanly, the outbreak is declared over, as WHO and Africa CDC did for Uganda in August 2026.

Why did ring vaccination work in DRC?

It concentrated a fast-acting vaccine on the people most likely to be infected next: contacts of each case and their contacts. In DR Congo's 2018-2020 outbreak more than 303,000 people were vaccinated this way. The approach depends on a vaccine matched to the species, which Bundibugyo lacks.