Euthanasia Statistics 2026: Global Data, Trends and Who Asks

In September 2018, Audrey Parker spoke to CBC News at her home city of Halifax, Canada. She was 57. She had stage 4 breast cancer that had spread to her bones and her brain. She was receiving palliative care, which is care that eases symptoms rather than trying to cure (CBC News, 2018).

"I had two years trying to get on top of the pain, and man, I suffered," she said.

Audrey had been approved for medical assistance in dying, known in Canada as MAID. But the law then had a rule. A person had to be able to consent again on the day itself. Audrey feared the cancer in her brain would take away her ability to decide before then. If that happened, she could lose the choice she had already been granted. So she chose a date earlier than she wanted.

"I think once I've signed the papers and have agreed, it should stand," she said.

The same CBC article gave one number. By the end of 2017, there had been 3,714 medically assisted deaths in Canada. Audrey became one more line in that count. This page is about lines like hers.

Statistics can feel cold. But each number in an annual report is a person, with a family, an illness and a reason. Dame Cicely Saunders, who founded the modern hospice movement, put it this way: "You matter because you are you, and you matter to the end of your life" (King's College London).

This guide gives you the most recent official numbers, checked on 11 October 2026. It also explains what the best research shows and where the data runs out. We use three labels to show how firm each finding is:

• Strong: many studies or full official records agree.
• Mixed: good studies disagree, or the data can be read in different ways.
• Early: only a few years of data, or only one study.

We do not take sides. Supporters and critics both use these numbers. Our aim is to help you read them clearly.

Assisted dying is still a small share of all deaths. But in some countries it is no longer rare. A 2025 review in The Lancet Public Health found that more than 200 million people live in places that allow some form of it (Sillitti et al., 2025). For the full list of places, see where euthanasia is legal in 2026.

Here are the latest official figures:

Netherlands. The Regional Euthanasia Review Committees (RTE) received 10,341 reports in 2025. That was 3.8% more than in 2024, and 6% of all deaths (RTE, 2026). In 2024 there were 9,958 reports, 10% more than the year before. That was 5.8% of deaths, up from 5.4% in 2023 (Lewitzka et al., 2026).

Belgium. The Federal Commission for the Control and Evaluation of Euthanasia (FCEE) recorded 4,486 euthanasia deaths in 2025 (4,487 in its later full report, the figure in the chart above). That was 12.4% more than the 3,991 in 2024. The share of deaths rose from 3.6% to 4%, or about 1 in 25 (RTBF).

Canada. In 2024, 16,499 people received MAID, 5.1% of all deaths and 6.9% more than in 2023. Since 2016 there have been 76,475 MAID deaths (Health Canada, 2025). Quebec alone made up 36.4% of all cases in 2024. Canada has the largest number of assisted deaths of any country.

Oregon. Oregon has one of the longest official records, with a report every year since 1998. In 2025, doctors wrote 637 prescriptions, and 400 people died after taking the medicine (Oregon Health Authority, 2026). That was fewer than the 421 deaths in 2024, but the state says the 2025 count will likely rise as late death certificates arrive.

Spain. In 2024, 426 people received help to die. Since the law began in June 2021, there have been 2,432 requests, and 1,123 (46%) ended with assisted death (Ministry of Health, Spain).

Switzerland. The Federal Statistical Office recorded 1,729 assisted suicides among residents in 2023. In the same year, there were 995 other suicides (Swiss Parliament, item 25.3148).

Germany. There is no national register. A 2025 study of death records in Bavaria found 316 assisted suicides between 2020 and 2023. By 2023 they made up about 1 in 10 suicides in Bavaria and 0.1% of all deaths (Buschner et al., 2025).

Other places. New Zealand, Australia, Colombia, Ecuador and Uruguay publish less data, or only a few years of it. Delaware, New York and Illinois began in 2026 and have no annual reports yet. France's law is expected to start in 2027. Our legal landscape page lists every law.

How big is 6% of deaths? Picture a Dutch village where 100 people die in a year. On average, six of them die by euthanasia. In Oregon, it would be one.

Strong evidence: numbers rise after a law starts. A landmark review in JAMA by Ezekiel Emanuel and colleagues looked at data from 1947 to 2016. It found that assisted deaths made up 0.3% to 4.6% of all deaths where legal, and that "the frequency of these deaths increased after legalization" (Emanuel et al., JAMA 2016). Today, the Netherlands and Canada are both above that old upper limit.

The long-term trends are clear:

• Netherlands: 1,933 reports in 2005, three years after the law began (Onwuteaka-Philipsen et al., Lancet 2012). 10,341 in 2025. That is more than five times as many.
• Canada: 3.3% of deaths in 2021, 4.1% in 2022 (Health Canada, 2023), 4.7% in 2023 and 5.1% in 2024. Health Canada first reported that growth averaged about 31% a year from 2019 to 2022, then slowed to 15.8% in 2023 (Health Canada, 2024). Its 2024 report revised the figures (the ones in the chart above), which puts growth in 2023 at 16.9%. In 2017, the total since 2016 was 3,714. By 2024, it was 76,475.
• Belgium: up 16.6% in 2024 (FPS Public Health) and 12.4% in 2025.
• Oregon: prescriptions rose about 5% in 2025, from 609 to 637. In 2023 they jumped about 29%, after the state dropped its rule that patients must live in Oregon.
• Bavaria: the rate rose from 0.1 to 1.3 per 100,000 people between 2020 and 2023.

Think of a new bridge over a river. At first, few people use it. Then more people learn it is there, and traffic grows. Over time, it may level off. With assisted dying, growth has slowed in some years and places. But it has not clearly stopped anywhere with a large program.

Not every number goes up. In the Netherlands, cases based mainly on psychiatric suffering fell by about 20% in 2025, to 174. And in Canada, yearly growth slowed from about 31% to 6.9% in 2024.

Why does it matter? Supporters see growth as people learning about a legal right. Critics see it as a sign that the practice is becoming normal, and that rules slowly widen. Both can be true at once. The numbers alone cannot tell you which story is right.

Strong evidence: most people are older. In Canada, the median age in 2024 was 77.9 (Health Canada). In Oregon, 88% were 65 or older. In Belgium, 73.7% were over 70 and 45% were over 80. Only 1.4% were under 40. One was a minor, the seventh since Belgium allowed euthanasia for minors in 2014 (RTBF). In Spain, 75.89% of completed requests in 2024 came from people over 60.

Sex: roughly equal overall, with some differences. In Canada, 51.8% were men in 2024. In Belgium, 52.2% were women in 2025. A 2026 study of 25 years of Oregon data found women were about half of all deaths (McIntosh and Canetto, 2026). But some groups differ:

• On Canada's track for people who are not near death, 56.7% were women.
• In Bavaria, almost two-thirds of assisted suicides were women.

This is very different from suicide in general, where men are the large majority. The two are different events, with different causes.

Education and race. The Emanuel review found that typical patients are "older, white, and well-educated". In Oregon in 2025, 94% were white and half had a bachelor's degree or higher. In Canada, 95.6% of those who answered the question were white. A 2007 study said Oregon and Dutch patients "appeared to enjoy comparative social, economic, educational, professional and other privileges" (Battin et al., 2007).

Why might this be? There are several possible reasons. People with more education may find it easier to navigate the system. Some minority communities have strong reasons, based on history, to distrust medicine. Religion and culture also play a part. Critics also ask whether some groups face barriers to access.

Culture inside one country. Belgium shows how much culture matters. Requests in Dutch were about three times as common as requests in French in 2025. The commission's co-chair, Jacqueline Hermans, suggested that Flemish doctors talk more openly with patients about all the options, while French-speaking medicine is more paternalistic, meaning doctors decide more for the patient.

City and country. Data here is limited. Most reports do not break down cases by city and countryside, so we cannot draw firm conclusions.

Strong evidence: cancer is the most common illness everywhere. The Emanuel review found that more than 70% of cases involved cancer. Today the share is often lower, as other illnesses grow:

In Canada, people could have more than one condition, so the numbers add up to more than 100%.

Neurological diseases. These include ALS (a disease that slowly stops the muscles working), Parkinson's disease and multiple sclerosis. They take up a much larger share of cases than their share of deaths in general. In Spain they are now the most common reason. On Canada's track for people not near death, 51.6% had a neurological condition.

Several conditions at once. In Belgium, almost 3 in 10 people had multiple illnesses. In the Netherlands in 2025, there were 475 reports about people with an "accumulation of age-related conditions", such as poor sight, hearing loss and frailty.

Dementia. In the Netherlands, 499 people with dementia received euthanasia in 2025. In 11 cases, the person could no longer decide about their request at the time.

Psychiatric conditions. These are allowed only in a few countries, and they are the most debated. In the Netherlands, 174 cases in 2025 involved mainly psychiatric suffering, about 1.7% of the total. None involved a minor. In Canada, people whose only condition is a mental illness are not eligible. That exclusion was due to end in March 2027. On 7 October 2026, Justice Minister Sean Fraser said a new bill this fall will extend it with no end date. He cited weak access to mental health care in some communities, and no agreement on when a mental illness is incurable (CBC News). Read more in MAID for mental illness: why Canada said no.

Many people assume pain is the main reason. The data says otherwise.

Strong evidence: loss of independence matters more than pain. In Oregon in 2025, the three most common end-of-life concerns were:

• loss of autonomy, meaning control over one's own life (89%)
• being less able to do things that made life enjoyable (89%)
• loss of dignity (65%)

Concern about pain was reported by 37%, according to a summary of the same report by Americans United for Life, a group that opposes these laws.

Canada's 2024 data shows the same pattern. On the main track, 95.1% reported loss of ability to take part in meaningful activities. Then came loss of ability to do daily tasks (85.4%) and loss of independence (75.4%). Pain, or worry about pain, was reported by 55.9%.

Two figures deserve close attention.

• Feeling like a burden. About half of Canadians who received MAID (48.4% on the main track, 50.3% on the other) said they felt a burden on family, friends or carers. The 2026 Oregon study found burden motives rising over time, mainly among men and older adults (McIntosh and Canetto, 2026).
• Loneliness. On Canada's main track, 21.9% reported isolation or loneliness. On the track for people not near death, it was 44.7%, about twice as high.

There is an important gap here. In most places, the reasons are written down by the doctor, not by the patient. The Oregon study notes that the record "has no information about motives directly from petitioners". So we know what doctors heard, or believed. We do not always know what patients felt.

Mixed evidence: depression. In 2008, Linda Ganzini and colleagues interviewed 58 Oregonians who had asked about aid in dying. Fifteen met the criteria for depression. Most people who received a prescription were not depressed, but three were (Ganzini et al., BMJ 2008). The authors warned that the law "may fail to protect some patients whose choices are influenced by depression". In 2025, only two Oregon patients were referred for a psychiatric evaluation, according to the AUL summary.

Depression is common in serious illness, and it can be treated. Assisted dying laws are written for physical illness under strict rules. They are not an answer to depression, loneliness or distress. If you are struggling, please reach out today. Our resources page lists people who can help.

Not always. And not everyone who asks is approved.

Oregon: a safety net that many never use. In 2025, 637 people received a prescription. According to the AUL summary of the state report, 358 of them were known to have taken it, 100 died without using it, and the status of 179 was unknown. Supporters often say that just having the prescription brings peace of mind. For them, it works like a fire extinguisher on the wall. You may never use it, but knowing it is there changes how you feel.

Oregon also shows that predictions about time are uncertain. In 2025, 24 of the 400 people who died (6%) had lived longer than six months after receiving their prescription.

Canada: what happened to 22,535 requests. In 2024, Canadian practitioners received 22,535 written requests. This is what happened:

Of those who died first, 1,661 had never chosen a date. Another 1,502 had been approved but died before the date came. For 107 people, the reason was that they lost the ability to consent before MAID, without a written waiver in place. That is Audrey Parker's fear, in a single number.

Spain: long waits. Of 929 requests completed in 2024, 45.86% ended in assisted death. A third (33.15%, or 308 people) ended because the person died while the request was still being processed. Another 15.18% were refused, and 5.81% were withdrawn by the person.

The Spanish report adds a hard detail. For those who died during the process, the median time from request to death was 27 days. That is shorter than the minimum time the law's own steps require. The median wait from request to decision was 41 days (Ministry of Health, Spain).

What do these numbers mean? Supporters point to the deaths during the wait as a sign the process is too slow. Critics point to the refusals and withdrawals as proof that checks matter. Both are reading the same table.

Strong evidence: most people in Western countries support legal assisted dying, at least for the terminally ill. But support depends a lot on how the question is asked.

United States. In Gallup's 2024 poll, 71% said doctors should be allowed by law to end a patient's life "by some painless means" if the patient and family ask. A second question asked whether doctors should be allowed to "assist the patient to commit suicide" if they are terminally ill and in severe pain. Support was 66%. Gallup first asked in 1947, when only 37% agreed (Gallup, 2024).

Legal and moral are not the same. Only 53% said doctor-assisted suicide was morally acceptable, and 40% said it was wrong. By 2026, Gallup found views "nearly evenly split" (Gallup, 2026). A 2025 Pew survey found that 63% did not think it was morally wrong: 34% said it was acceptable and 29% said it was not a moral issue (Pew Research Center, 2026).

Religion is the strongest divide. In Gallup's data, 67% of people who seldom or never attend religious services called it morally acceptable. Among weekly attendees, only 29% did. Pew found that 59% of US Catholics did not see it as morally wrong, even though Catholic teaching forbids it. Among white evangelical Protestants, 60% said it was wrong.

Politics. In the US, 79% of Democrats supported legal euthanasia in 2024, compared with 61% of Republicans.

Britain. In the 2025 British Social Attitudes survey, 79% said a doctor should "definitely" or "probably" be allowed to end the life of someone with a terminal, painful illness who asks. For someone with a painful illness who will not die, support was 62%, up from 41% in 1995 (NatCen, 2026).

Polls do not make laws. Despite that support, MPs rejected a bill for England and Wales by 286 votes to 270 in September 2026 (The Guardian). An earlier bill had passed the Commons in June 2025, but ran out of time in the House of Lords in April 2026. Three Labour MPs who voted against the new bill said it was "unsafe and unworkable", not that they opposed the idea in principle.

Elsewhere. The Emanuel review found US support had stayed in a range of 47% to 69% since the 1990s, depending on wording. In Western Europe it was strong and rising. In Central and Eastern Europe, it was falling.

Here is a question for you. If a poll asks about "ending a life by painless means", and another asks about "assisted suicide", are they asking the same thing? Many people answer them differently.

This is the most important question in the whole debate. The honest answer is that the evidence is mixed.

Studies that found no extra risk. In 2007, Margaret Battin and colleagues checked Oregon and Dutch data for 10 groups. They included older people, women, uninsured and poor people, disabled people, minors and people with depression. They found no higher risk in any group except people with AIDS (Battin et al., 2007). The 2016 JAMA review agreed: "In no jurisdiction was there evidence that vulnerable patients have been receiving euthanasia or physician-assisted suicide at rates higher than those in the general population."

Studies that raised concerns. A survey of deaths in Flanders, Belgium, in 2007 found that life-ending drugs were used without an explicit request in 1.8% of deaths. Most of these patients were over 80 and in hospital. In 77.9% of these cases, the decision was not discussed with the patient (Chambaere et al., CMAJ 2010). The authors noted these cases happened in different groups and settings from euthanasia on request.

In the Netherlands, ending a life without an explicit request fell from 0.8% of deaths in 1990 to 0.2% in 2010. The share of euthanasia cases reported to review committees was 77% in 2010 (Onwuteaka-Philipsen et al., Lancet 2012). The authors said the law had made practice "relatively transparent". Critics point out that nearly a quarter of cases were still not reported.

Canada's newer track. Since 2021, people in Canada who are not near death can ask for MAID under stricter rules, called Track 2. In 2024, 732 people (4.4%) received MAID this way. Of them, 61.5% reported a disability.

Ontario's coroner studied these cases closely. In 2023, 116 of Ontario's 4,644 MAID deaths were on Track 2. These people were more often women (61%), more often younger, and more likely to live alone (Ontario MAiD Death Review Committee, 2024). One case was a man in his 40s with a bowel disease and a history of mental illness. His substance use was not explored, and the committee said he appeared "socially vulnerable and isolated".

Ontario's wider review is more reassuring in one way. No concerns were found in 3,861 of 4,644 MAID deaths in 2023. But 6% of cases led to formal notices about legal requirements (Ontario MAiD Review Team). Quebec's commission reported 23 cases of possible non-compliance in 2022–23 (Policy Options). In the Netherlands, review committees found that doctors failed the due care rules in 7 of 10,341 cases in 2025.

The disability rights view. In 2021, UN human rights experts said: "Disability should never be a ground or justification to end someone's life directly or indirectly" (EWTN News). The US group Not Dead Yet opposes these laws "as deadly forms of discrimination" (Not Dead Yet). Their concern is not only who uses the law. It is why, when support services are missing. In Canada in 2024, 32.5% of MAID recipients had received disability support services.

So which is it? Population studies look at large groups and find no clear pattern of harm. Case reviews look at single people and find worrying stories. Both kinds of evidence matter. A low average risk is little comfort to the one person who should have been protected.

Strong evidence: most people who choose assisted dying are already receiving end-of-life care. In Oregon in 2025, 92% were enrolled in hospice. In Canada in 2024, 74.1% had received palliative care. Only 2.5% of recipients needed palliative care but had not received it. For 91.2% of these, the services were available.

But there is a gap. On Canada's Track 2, only 23.2% had received palliative care. Many of these people were not dying, so palliative services may not have been designed for them.

What the research says about care at the end of life:

• The SUPPORT study (JAMA, 1995). This study followed 9,105 seriously ill patients in five US hospitals. Families reported moderate to severe pain at least half the time for 50% of conscious patients who died in hospital. A nurse-led plan to improve communication did not change these results (SUPPORT, 1995). It showed how poor end-of-life care could be.
• The bigger picture (JAMA, 2016). Dio Kavalieratos and colleagues pooled 43 trials with 12,731 patients. Palliative care improved quality of life and symptoms in the first few months. But it was not linked to longer survival (Kavalieratos et al., JAMA 2016). Strong for quality of life; not confirmed for survival.
• Early palliative care (NEJM, 2010). Jennifer Temel and colleagues tested early palliative care in 151 people with advanced lung cancer. Those who got it had better quality of life and less depression (16% vs 38%). They also lived longer, with a median of 11.6 months against 8.9 months (Temel et al., NEJM 2010). It was one trial, but it changed practice worldwide.
• Dignity therapy (Lancet Oncology, 2011). Harvey Chochinov's team tested a short talking therapy for 441 dying patients. It did not reduce distress or the wish to die. But patients said it helped them, and it improved their end-of-life experience (Chochinov et al., 2011). Mixed.

Some critics fear that assisted dying could take the place of good care. The Catholic Church warns of "vulnerable persons who needed only to be better cared for and comforted" (Samaritanus bonus, 2020). Supporters reply that the two can exist side by side, and that the data shows most patients already get care.

What about cost? In 2017, a study in the CMAJ estimated that MAID could reduce Canadian health spending by $34.7 million to $138.8 million a year (Trachtenberg and Manns, 2017). It was a model built on Dutch and Belgian data, not a count of real savings. The authors were clear: "We are not suggesting medical assistance in dying as a measure to cut costs." They added that patients and doctors should never consider cost. Critics still say any link between dying and saving money is dangerous. Learn more on our palliative care page.

Only a small share of doctors take part where assisted dying is legal. Most never do.

• United States: In Oregon and Washington, less than 1% of licensed doctors write a prescription each year (Emanuel et al., 2016).
• Netherlands: About 60% of Dutch doctors have granted a request at some point in their career, according to the same review.
• Canada: In 2024, 2,266 practitioners provided MAID. Nearly all (93.2%) were doctors, and 6.8% were nurse practitioners.

Doctors may refuse on grounds of conscience in every system covered here. This creates gaps in some regions. In Oregon, the median time patients had known their prescribing doctor was about four weeks in 2025, according to the AUL summary. Critics say this makes it harder for doctors to spot pressure or depression. Supporters say patients must often look beyond their regular doctor, who may object.

Some doctors help patients die even where the law forbids it. Measuring this is very hard.

Early evidence from anonymous surveys. In 1996, Diane Meier and colleagues surveyed 1,902 US doctors. Of all doctors in the sample, 18.3% had been asked by a patient for help with suicide. About 3.3% said they had written at least one prescription to hasten death, and 4.7% said they had carried out euthanasia at least once (Meier et al., NEJM 1998). All of this was illegal at the time.

In the UK, a 2007–2008 survey of doctors estimated that 0.21% of deaths involved voluntary euthanasia. About 0.30% involved ending a life without an explicit request (Seale, 2009). Other end-of-life decisions, such as stopping treatment, were rarely judged to shorten life by more than a day.

Why the numbers are uncertain. Doctors may not admit illegal acts, even in anonymous surveys. The meaning of "hastening death" varies. Strong pain relief at the end of life is legal, and its intent is rarely written down.

What this means for the debate. Supporters argue that legal rules bring hidden practice into the open, where it can be checked. The Dutch fall in unrequested cases after the law is often cited. Critics argue that hidden practice is rare and that legal change brings far larger numbers. The data is not good enough to settle this.

Strong evidence: the data is uneven. The Lancet Public Health review found that different terms, and the lack of a specific disease code for assisted dying, block fair comparison between countries (Sillitti et al., 2025).

Here are the main traps:

• Different words. In Canada, MAID covers both euthanasia and assisted suicide. In the US, it means only assisted suicide.
• Different counts. Oregon reports prescriptions and deaths. Canada reports provisions. Switzerland's figures cover residents only.
• Different rules. Countries that allow euthanasia by a doctor have much higher rates than those that allow only assisted suicide. Where both are allowed, euthanasia makes up most cases (Sillitti et al.).
• Who fills in the form. Reasons are mostly recorded by doctors. Unknowns can be large, as in Oregon's 179 prescriptions with no known outcome.
• Unchecked numbers. In Germany, figures from right-to-die organisations are published but not checked by any authority (Lewitzka et al., 2026).

When you see a euthanasia statistic in the news, ask five questions:

1. Is it a share of all deaths, or a raw number?
2. Does it include euthanasia, assisted suicide, or both?
3. Which year is it from, and is it final?
4. Who collected it, and who might benefit from how it is presented?
5. Whose voice is missing: the patient's, the family's, or the doctor's?

The same problems come up with other deaths that laws define in different ways. Deaths from overwork are one example: Japan, South Korea, China and Taiwan each count them by different rules. See Worked to death: overwork deaths around the world.

For real stories behind these numbers, see our case studies. For both sides of the debate, see arguments for and against.

Audrey Parker died at home in Halifax on 1 November 2018, the date she chose, earlier than she wanted. Three days before, she recorded a video. "People like me who have already been assessed and approved are dying earlier than necessary because of this poorly thought-out law," she said (CBC News, 2019).

Her case reached Parliament. Bill C-7, passed in March 2021, made a change for people like her. A person whose death is reasonably foreseeable, and who risks losing capacity, can now make a written arrangement in advance. Justice Canada said the proposal "would respond to circumstances like that of Audrey Parker" (Justice Canada). Many people call it "Audrey's Amendment".

Look again at the numbers. When Audrey spoke, the total was 3,714. By the end of 2024 it was 76,475. Each year Health Canada still counts people who lost the ability to consent before MAID, without a waiver in place. In 2024 there were 107 of them. The rule changed, but the problem has not fully gone away.

The debate Audrey started is still moving. When Canada's Justice Minister announced the mental illness exclusion in October 2026, he also said the new bill will "open the door" to advance consent. This would cover people whose incurable illness will later take away their capacity to decide (CBC News). Audrey's Amendment helps only people who are already approved and near death. The details of the new idea will be worked out by MPs in committee.

Canada's Supreme Court, in Carter v Canada (2015), had said the choice belongs to a competent adult who "clearly consents" (Carter). Audrey's story showed how hard that idea is to turn into rules and numbers. Supporters saw her as proof the law was too strict. Critics warn that each loosening of the rules makes the next one easier.

"I just hope that they make some good assessments that will help people in the future," Audrey told CBC.

That is what good statistics can do. They can show where care fails, where rules work and where they do not. But they can never show everything. Behind every percentage on this page is someone like Audrey. As Cicely Saunders taught, each of them mattered to the end of their life.

When you next see one of these numbers, what question will you ask about the person behind it?