SARS and MERS: Coronaviruses Explained

SARS and MERS are the two coronavirus diseases that showed epidemiologists what the family could do before 2020. SARS-CoV-1 killed 774 people in 2003 and was stopped without a vaccine. MERS-CoV has killed more than a third of the people it is reported to have infected, and it still jumps from camels to humans every year. Comparing the two explains why one outbreak ended, why the other persists, and why SARS-CoV-2 slipped past tools that had worked on its cousin.

Educational content, not medical advice. This article describes general, well-established findings from public health agencies and published studies. Talk to a doctor about any personal health concern; every person's situation differs.

What makes a coronavirus a coronavirus

Coronaviruses are RNA viruses whose outer envelope is studded with club-shaped spike proteins. Under an electron microscope the spikes look like the points of a crown, which gave the family its name. The family is divided into four genera: alpha, beta, gamma, and delta. Two of them contain the coronaviruses that infect people.

The spike is the part that matters most for this story. It works like a key cut to fit a particular lock, the receptor protein on the surface of a host cell. Once the key fits, the viral and cellular membranes fuse and the viral genome slips inside. Which receptor a spike recognizes decides which tissues the virus can infect, how well it spreads, and which animals can carry it. For the general mechanism of how viruses enter and take over cells, see how viruses work, from infection to replication.

Seven coronaviruses are currently known to infect humans. Four of them are ordinary, and three are the outbreak viruses.

Virus	Genus	Typical impact
HCoV-229E	Alphacoronavirus	Common cold
HCoV-NL63	Alphacoronavirus	Common cold
HCoV-OC43	Betacoronavirus	Common cold
HCoV-HKU1	Betacoronavirus	Common cold
SARS-CoV-1	Betacoronavirus	Severe pneumonia, 2002 to 2003 outbreak
MERS-CoV	Betacoronavirus	Severe pneumonia, ongoing sporadic cases since 2012
SARS-CoV-2	Betacoronavirus	COVID-19 pandemic from 2019

The four common-cold viruses are covered in seasonal coronaviruses OC43, 229E, NL63 and HKU1. The names SARS-CoV-1 and SARS-CoV-2 imply that they are the same virus, and the truth is more interesting. The International Committee on Taxonomy of Viruses placed SARS-CoV-2 in the same species as the original SARS virus, as a sister clade to it, which is why it was named the way it was (Coronaviridae Study Group, Nature Microbiology 2020). Sharing a species did not mean sharing behavior, as the later sections show. MERS-CoV is a separate betacoronavirus species.

How SARS emerged in 2002 and 2003

WHO's final tally places the earliest onset of a probable SARS case on November 16, 2002, in China (WHO summary of probable SARS cases). The outbreak began in China and spread to 28 other countries, and the causative agent was identified as a coronavirus at the end of February 2003 (WHO SARS topic page). A German-led team reported the new virus in NEJM, noting that it was only distantly related to known coronaviruses, sharing roughly 50 to 60% of its nucleotide sequence with them (Drosten et al., 2003).

The receptor: ACE2

Within months, researchers showed that the SARS spike binds a human protein called angiotensin-converting enzyme 2 (ACE2). In cell experiments, soluble ACE2 blocked the spike from binding, and cells carrying ACE2 supported more efficient viral replication (Li et al., Nature 2003). ACE2 is an ordinary enzyme that cells make for their own reasons, so the virus is borrowing a doorknob that was already on the cell.

The animal route: from bats to civets to people

Tracing where SARS came from took years and three kinds of evidence. First, in 2003 researchers isolated SARS-like viruses from Himalayan palm civets at a live-animal market in Guangdong, and found evidence of infection in a raccoon dog and in people who worked at the same market. All the animal isolates kept a stretch of 29 nucleotides that most human isolates had lost, which suggested the virus had changed as it crossed species through the market (Guan et al., Science 2003). At that point the natural reservoir was still unknown.

Second, in 2005 a team reported that bats are natural hosts of SARS-related coronaviruses, and that these viruses vary more genetically than SARS isolates from humans or civets (Li et al., Science 2005). Third, in 2013 researchers isolated a live SARS-like virus (WIV1) from horseshoe bat feces in Yunnan that could enter cells using the ACE2 of humans, civets, and horseshoe bats, and concluded that some bat viruses may infect people without an intermediate host (Ge et al., Nature 2013).

Picture the route as stepping stones across a river. Bats are the far bank, where the viruses live harmlessly and diversify. Markets that hold wild animals in close quarters supply the middle stones, where a virus gets many chances to practice on new hosts. A human is the near bank. Most hops fail. The ones that succeed are the ones where the spike happens to fit the human lock. For the general pattern, see how viruses jump from animals to humans.

How SARS spread, and why hospitals were the engine

SARS spreads through small respiratory particles in the air, and also indirectly when infectious virus lands on surfaces others touch. International air travel carried it across borders. The incubation period was usually 2 to 7 days and up to 10, and in 10 to 20% of cases the illness was severe enough to need intubation and mechanical ventilation (WHO).

A mathematical analysis in Science estimated that, without control measures, a single infectious SARS case would infect about three others (Lipsitch et al., 2003). That number is high enough to grow an epidemic and low enough to stop one with good control, which is exactly what happened.

Healthcare workers as a warning sign

SARS patients were often treated before anyone knew what they had. WHO's final table lists 1,706 healthcare workers among the probable cases, about 21% of the total (calculated from the table's totals). Hospitals concentrate sick people, procedures that make aerosols, and staff moving between rooms, so a respiratory virus can use them as amplifiers.

Amoy Gardens

A residential complex in Hong Kong provided the clearest example of how environment can matter. An NEJM analysis of the first 187 SARS cases at Amoy Gardens found that 99 patients lived in Building E, which also housed the index patient, while only five lived outside the seven buildings labeled A to G. The paper's title says what the authors concluded: it presented evidence of airborne transmission (Yu et al., NEJM 2004). The concentration in one building pointed to a shared environmental route rather than person-to-person contact alone.

The final numbers for SARS

WHO's tally as of December 31, 2003 recorded the following.

Measure	Value
Probable cases	8,096
Deaths	774
Case fatality ratio	9.6%
Areas reporting cases	29 (China, Hong Kong, Macao, and Taiwan are listed separately)
Healthcare workers affected	1,706 (about 21%)
Earliest onset of a probable case	November 16, 2002

Source: WHO summary of probable SARS cases. Figures commonly quoted elsewhere, such as 8,098 cases, differ by a few cases because tallies were revised after 2003.

A case fatality ratio is the share of diagnosed patients who died, and it depends heavily on who gets counted. Early in an outbreak, mild cases are missed, which inflates the ratio. Late in an outbreak, some patients are still sick, which deflates it. The 9.6% figure is a final, retrospective count in a population that was mostly tested in hospitals, so it should be read as a measure of severe disease, not as the risk for everyone infected.

How SARS was stopped without a vaccine

Containment came from old-fashioned public health. WHO lists early detection, isolation of suspected and probable cases, contact tracing, quarantine of suspected contacts for 10 days, exit screening at airports, and disinfection of aircraft and cruise ships with SARS cases on board (WHO).

Those measures worked because of when SARS patients were infectious. In a prospective study of a community outbreak, viral load in nasopharyngeal samples from 14 patients peaked at about day 10 of illness, well after symptoms began, and by day 15 it had fallen below the level at admission (Peiris et al., Lancet 2003). Peak shedding arriving after fever and cough meant the sick announced themselves before they were at their most contagious. Isolating people with symptoms cut off most of the chain.

Imagine a fire alarm that sounds before the fire reaches its full size. A building can be evacuated in time. That is why a disease with a reproduction number around three could be wrestled to a stop.

The last cases

A few sporadic cases followed. Four SARS patients were identified in Guangdong in the winter of 2003 to 2004, and two were linked to a Guangzhou restaurant that served palm civets (Emerging Infectious Diseases 2005). That episode showed the animal route was still active while the human chains had been broken.

MERS: the virus that never left

In 2012, doctors in Saudi Arabia treated a 60-year-old man with acute pneumonia who then developed renal failure and died. From his sputum, researchers isolated a previously unknown coronavirus, a new betacoronavirus species whose closest known relatives were the bat coronaviruses HKU4 and HKU5. The authors noted that his illness closely resembled SARS (Zaki et al., NEJM 2012). It was later named Middle East respiratory syndrome coronavirus, MERS-CoV.

A different receptor

MERS-CoV does not use ACE2. It binds a protein called dipeptidyl peptidase 4 (DPP4, also known as CD26). In 2013, researchers showed that antibodies against DPP4 blocked infection of human airway cells, and that expressing human or bat DPP4 made otherwise resistant cells susceptible. They suggested that DPP4's conservation across species hints at the virus's potential host range (Raj et al., Nature 2013).

The camel connection

WHO identifies dromedary camels as the primary reservoir of MERS-CoV. Infected camels do not get sick, so detection depends on laboratory testing. People can be infected through direct contact with dromedaries, and possibly indirectly through camel-related raw products such as milk and urine (WHO MERS fact sheet).

A 2014 NEJM report showed how direct this can be. A man died of laboratory-confirmed MERS after close contact with camels that had runny noses. Virus was isolated from one of his nine camels, serology indicated it had been circulating in the herd before his illness, and the full genome sequences from man and camel were identical (Azhar et al., NEJM 2014).

Antibodies suggest the virus is widespread in camels. A 2026 meta-analysis of studies in African dromedaries reported a pooled antibody prevalence of 75.1%, with 78.5% in adults and 42.5% in juveniles, and viral RNA in 6.2% (Veterinary Research Communications 2026). Antibodies show past infection, not current infectiousness, so the viral RNA figure is the more cautious one.

Where do bats fit? The earliest sequence data tied MERS-CoV's nearest relatives to bats, as noted above, so its deeper ancestry plausibly lies there, as with SARS. The virus that matters to people today, though, circulates in camels.

How lethal is MERS, really?

WHO reports more than 2,600 laboratory-confirmed MERS cases from 27 countries, with about 37% of reported cases having died, a figure WHO itself says is likely an overestimation (WHO fact sheet, December 11, 2025). Why would the official number be too high?

The answer lies in who gets counted. In a 2013 analysis, researchers estimated that 111 confirmed or probable cases corresponded to about 940 symptomatic cases (95% confidence interval 290 to 2,200), meaning at least 62% of symptomatic cases went undetected. Among primary cases picked up by routine surveillance the apparent case fatality was 74%, but among secondary cases, found by tracing contacts of known patients, it was 20% (Cauchemez et al., Lancet Infectious Diseases 2014). Severe cases get hospitalized and tested. Mild ones do not. That is a bit like estimating how dangerous a hiking trail is by counting only the people carried out by helicopter.

Even so, MERS remains one of the deadliest human coronavirus infections. WHO lists people over 65 and those with weakened immune systems or chronic conditions, including cancer, lung disease, cardiovascular disease, and diabetes, as higher-risk groups.

Why MERS has not gone pandemic

The same analysis estimated that, without control measures, the reproduction number of MERS was between 0.8 and 1.3, and the upper bound from three independent data sources was 1.2 to 1.5. A reproduction number below one means chains fizzle out on their own. WHO states that human-to-human transmission is possible mainly among close contacts and in healthcare settings, has been limited elsewhere, and that no sustained human-to-human transmission has been documented anywhere.

That profile has a consequence. MERS does not grow exponentially in communities, but it can erupt in hospitals, and the camel reservoir keeps resupplying it. The disease is persistent, not explosive.

The 2015 outbreak in South Korea

A single traveler returning from the Middle East in May 2015 carried MERS into the Korean health system. WHO records 186 infections and 38 deaths from that outbreak. A later review of the Korean experience found that nosocomial (hospital-acquired) transmission occurred in 16 hospitals, that 44% of the 186 cases were nosocomial, and that five superspreaders were responsible for 83% of transmission events. Authorities quarantined 16,993 people for 14 days (Oh et al., Korean Journal of Internal Medicine 2018).

An early assessment during the outbreak estimated a mean incubation period of 6.7 days and a mean serial interval, the gap between successive cases in a chain, of 12.6 days (Eurosurveillance 2015).

Superspreading is the key idea. Most infected people pass the virus to no one or to very few. A small number pass it to many, and they account for most of the spread. It resembles a forest fire season in which a handful of sparks burn most of the acreage. In Korea the sparks landed in crowded emergency rooms and shared wards, where people with undiagnosed infections interacted with many patients and visitors.

The lesson is that a virus with a low reproduction number can still do great damage when infection control fails in the right place. Close attention to hospital procedures matters as much for MERS as any antiviral does.

SARS, MERS, and SARS-CoV-2 side by side

Feature	SARS-CoV-1	MERS-CoV	SARS-CoV-2
First recognized	November 2002 (earliest probable onset)	2012, Saudi Arabia	Late 2019
Receptor	ACE2	DPP4 (CD26)	ACE2
Animal link	Bats; civets at live-animal markets	Dromedary camels; distant bat relatives	Not established
Reported cases	8,096 probable	More than 2,600 confirmed	Nearly 780 million reported to WHO
Reported deaths	774	About 37% of reported cases	More than 7.1 million reported to WHO
Crude fatality (reported deaths divided by reported cases)	9.6%	About 37%, likely overestimated	About 0.9%, heavily dependent on testing
Incubation	Usually 2 to 7 days, up to 10	Mean 6.7 days in the Korean outbreak	Usually 3 to 6 days
When infectious	Mostly after symptoms began	Mostly in close contact and healthcare settings	A large share before symptoms
Spread	Contained in 2003	Limited human chains; camel spillover continues	Pandemic

Sources: WHO fact sheets and the studies cited above. The SARS-CoV-2 figures come from WHO's COVID-19 fact sheet, which notes that actual numbers are likely higher than those reported. The crude fatality for SARS-CoV-2 is simple division of the two reported totals and is not an infection fatality rate.

Why timing decided everything

SARS-CoV-2 flipped the relationship between symptoms and spread. A modeling study of viral shedding and transmissibility estimated that 44% of secondary infections (95% confidence interval 25 to 69%) were infected during the index case's presymptomatic stage (He et al., Nature Medicine 2020). WHO says that infected people may also spread the virus without symptoms or in the days before symptoms develop (WHO COVID-19 fact sheet). Usual onset is 3 to 6 days after exposure.

Return to the fire alarm. With SARS, the alarm rang before the fire peaked. With SARS-CoV-2, the fire spread through a good share of the building before anyone smelled smoke. Isolating symptomatic people still helps, but it cannot by itself stop chains that run partly through people who feel well. Other tools, such as testing, contact tracing, and vaccines, had to carry more of the weight.

MERS sits in a third position. It spreads mainly among close contacts and in healthcare settings, and chains rarely sustain themselves in communities. The biology that makes each virus controllable or not comes down to three variables: when people are infectious, how many people each case infects, and how often the virus re-enters from animals.

The way the immune system responds to a new virus, including how antibodies and T cells build up over days, is laid out in how the human immune system fights viruses step by step.

What the outbreaks taught, and what was missed

Some lessons are visible in the numbers above. Hospitals were amplifiers in both outbreaks: about 21% of SARS cases were healthcare workers, and 44% of Korean MERS cases were hospital-acquired. That points to infection prevention in emergency rooms and wards, and to early recognition of respiratory illness, as basic safeguards.

Animal-human interfaces matter. The 2003 civet market and the camel herds of the Arabian Peninsula are interfaces where a virus gets repeated chances to infect people. Surveillance there, in animals and in people who work with them, is a cheaper form of defense than waiting for human cases.

Vaccines and treatments for MERS remain unfinished. WHO states that there is no licensed vaccine or specific treatment, that care is supportive, and that several candidate vaccines and treatments are in development, with some plans for Phase II trials. The most widely recognized vaccine success of the pandemic era is described in mRNA vaccines and the COVID-19 breakthrough. Surveillance of other animal viruses with pandemic potential follows the same logic, as seen with H5N1 bird flu, and the way a different respiratory virus behaves is covered in how seasonal flu works.

Researchers also use computational methods to flag which animal viruses may be primed to jump, a field described in AI and viral evolution prediction. For a broader look at what makes certain pathogens so dangerous, see the deadliest viruses on the planet.

What people get wrong about SARS and MERS

"The higher the fatality rate, the greater the global threat." Fatality and spread are separate dials. MERS is far deadlier per reported case than COVID-19, yet its reproduction number sits near or below one in community settings, so it has not caused a pandemic. A virus with a lower fatality rate that spreads easily and silently can kill more people in total, as the totals in the table show.

"SARS faded because it weakened." There is no evidence for that in the sources above. It ended because isolation, quarantine, and contact tracing broke the chains while the virus was still dangerous.

"A reported fatality rate is the chance of dying if infected." It is the share of counted cases who died. Because mild infections are missed, it overstates risk for the average infected person, as the MERS analysis showed.

"SARS-CoV-1 and SARS-CoV-2 are the same thing." They share a species classification and a receptor, but they differ in timing of infectiousness and in spread. The reverse mistake, treating them as unrelated, is equally wrong.

"Bats infect people directly." Sometimes that may be possible, as the WIV1 findings suggest, but in the documented cases of SARS and MERS, other animals sat between bats and people. The route varies by virus.

Current risk

As of the sources cited, SARS-CoV-1 is not a known current threat to people, since its human chains were broken in 2003, although related viruses persist in bats. MERS-CoV remains an ongoing but geographically focused concern. Cases continue to be reported, mainly linked to dromedary camels and healthcare settings, and the risk to the general public is limited by its poor person-to-person spread. WHO identifies direct contact with dromedaries, and possibly raw camel products, as the exposure routes.

Both viruses are reminders that the coronavirus family has produced three novel human pathogens in under two decades, a point the ICTV authors make in describing SARS-CoV-2 as the third major epidemic in about two decades caused by an animal coronavirus jumping to humans. Machine-learning forecasting of outbreaks is covered in machine learning and pandemic prediction.

Frequently asked questions

Q: Why did SARS disappear but MERS did not? A: SARS patients were mostly infectious after symptoms began, so isolation, contact tracing, and quarantine broke the chains in 2003. MERS-CoV circulates in dromedary camels, which are widespread and show no disease, so spillover into people continues.

Q: What is the difference between SARS and COVID-19? A: Both are caused by betacoronaviruses in the same species that use the ACE2 receptor. SARS-CoV-1 spread mainly after symptoms appeared, which made isolation effective. SARS-CoV-2 spreads substantially before symptoms, with one study estimating 44% of secondary infections occurred during the index case's presymptomatic stage, and it infected hundreds of millions of people.

Q: How deadly is MERS compared with other viruses? A: WHO reports about 37% of reported cases have died, but it calls that likely an overestimation. An analysis of early data estimated at least 62% of symptomatic cases were missed, so the true risk for an average infected person is lower than the headline figure.

Q: Can you catch MERS from a camel? A: Yes. WHO says people can be infected through direct contact with dromedaries, and possibly indirectly through raw products such as milk and urine. A 2014 NEJM report documented a fatal case in which the man and one of his camels carried identical viral genomes.

Q: Is there a MERS vaccine? A: Not a licensed one. WHO states that there is no licensed vaccine or specific treatment, that care is supportive, and that several candidates are in development.

Q: Where did SARS come from? A: Evidence points to horseshoe bats as the natural reservoir. Palm civets at a Guangdong live-animal market carried closely related viruses in 2003, and a 2013 study found a bat virus that can use human ACE2. The exact chain of infection for the first human cases is not fully settled.

Q: How many people did SARS infect and kill? A: WHO's final tally, as of December 31, 2003, was 8,096 probable cases and 774 deaths across 29 areas, for a case fatality ratio of 9.6%.

The takeaway

SARS and MERS are often remembered as scary numbers, but their real value is as a pair of experiments in what makes an outbreak controllable. SARS was deadly and fast but announced itself with symptoms before peak infectiousness, so ordinary isolation was enough. MERS is deadlier per reported case but rarely sustains chains in the community, and it keeps returning because camels keep supplying it. SARS-CoV-2 combined a receptor shared with SARS, easy spread, and infectiousness before symptoms, and no earlier playbook fit it cleanly. The next coronavirus will not match any of these three, which is why watching animal reservoirs and hospital infection control matters before a case count rises.

Sources

• World Health Organization, Middle East respiratory syndrome coronavirus fact sheet (December 11, 2025).
• World Health Organization, Summary of probable SARS cases, November 1, 2002 to July 31, 2003.
• World Health Organization, Severe acute respiratory syndrome topic page.
• World Health Organization, Coronavirus disease (COVID-19) fact sheet.
• Zaki AM, et al. Isolation of a novel coronavirus from a man with pneumonia in Saudi Arabia. NEJM 2012. PMID 23075143.
• Raj VS, et al. Dipeptidyl peptidase 4 is a functional receptor for the emerging human coronavirus-EMC. Nature 2013. PMID 23486063.
• Li W, et al. Angiotensin-converting enzyme 2 is a functional receptor for the SARS coronavirus. Nature 2003. PMID 14647384.
• Drosten C, et al. Identification of a novel coronavirus in patients with severe acute respiratory syndrome. NEJM 2003. PMID 12690091.
• Guan Y, et al. Isolation and characterization of viruses related to the SARS coronavirus from animals in southern China. Science 2003. PMID 12958366.
• Li W, et al. Bats are natural reservoirs of SARS-like coronaviruses. Science 2005. PMID 16195424.
• Ge XY, et al. Isolation and characterization of a bat SARS-like coronavirus that uses the ACE2 receptor. Nature 2013. PMID 24172901.
• Azhar EI, et al. Evidence for camel-to-human transmission of MERS coronavirus. NEJM 2014. PMID 24896817.
• Cauchemez S, et al. MERS coronavirus: quantification of the extent of the epidemic, surveillance biases, and transmissibility. Lancet Infect Dis 2014. PMID 24239323.
• Oh MD, et al. Middle East respiratory syndrome: what we learned from the 2015 outbreak in the Republic of Korea. Korean J Intern Med 2018. PMID 29506344.
• Preliminary epidemiological assessment of MERS-CoV outbreak in South Korea, May to June 2015. Euro Surveill 2015. PMID 26132767.
• Yu ITS, et al. Evidence of airborne transmission of the severe acute respiratory syndrome virus. NEJM 2004. PMID 15102999.
• Lipsitch M, et al. Transmission dynamics and control of severe acute respiratory syndrome. Science 2003. PMID 12766207.
• Peiris JSM, et al. Clinical progression and viral load in a community outbreak of coronavirus-associated SARS pneumonia. Lancet 2003. PMID 12781535.
• He X, et al. Temporal dynamics in viral shedding and transmissibility of COVID-19. Nat Med 2020. PMID 32296168.
• Coronaviridae Study Group of the ICTV. The species Severe acute respiratory syndrome-related coronavirus: classifying 2019-nCoV and naming it SARS-CoV-2. Nat Microbiol 2020. PMID 32123347.
• SARS-CoV infection in a restaurant from palm civet. Emerg Infect Dis 2005. PMID 16485471.
• MERS-CoV serological and virological evidence in dromedary camels in Africa: systematic review and meta-analysis. Vet Res Commun 2026. PMID 41729334.

---

Read next:

• Seasonal Coronaviruses: OC43, 229E, NL63 & HKU1
• H5N1 Bird Flu Pandemic Risk: 2026 Update
• Machine Learning and Pandemic Prediction: Can AI Prevent the Next Outbreak?