Best Countries for Medical Tourism: 2026 Index

Visa-free counts will not tell you where to get a hip replaced next month. A survey of Americans who have never sat in that waiting room will not either.

The CitizenX Medical Tourism Index scores 21 countries on a narrower question. If you can pick a jurisdiction, where is private care usable: price, quality, access, and a legal right to stay when the treatment is not a one-week trip.

CitizenX is a technology service, not a law firm, and does not give legal, tax, medical, or accounting advice.

The ranking

We scored 21 countries across six pillars. Weights and the rubric are below. The workbook has the same numbers as formulas if you want to move a weight.

A high score is not a medical recommendation. A low score does not insult the doctors who work there. It means the country is a weaker option for a non-resident who may also need the right to remain.

Why this index exists

The usual rankings answer a different buyer.

The last full Medical Tourism Index came out in 2020–21. Researchers asked 3,030 Americans how they perceive 46 destinations. Canada and Singapore won. Countries that treat millions of foreign patients sat lower because the method was a brand survey.

Travel lists for 2026 put Türkiye, Thailand, India, Mexico, and South Korea first because those countries process the most foreign patients. Thailand reported about 3.07 million foreign patient visits in 2025. South Korea’s health-industry institute counted 2.01 million foreign patients in 2025, up 71.9% from the year before. Türkiye’s official 2024 figure was 1.51 million healthcare visitors and about $3.02 billion in spend. Volume in 2025 was lower, near 1.4 million. Those numbers are real. They measure throughput. Tunisia and Iran also post large "health tourist" counts. That still does not tell a family in California or Zurich whether to book the flight.

CitizenX clients are allocating jurisdiction. Healthcare is one of the services a country either provides or does not.

Methodology

Edition: August 2026. Twenty-one countries. Each pillar is 0–100. The composite is a weighted sum. Two people applying the same rubric to the same public sources should land within five points on a pillar.

The job is to score a country as a healthcare option for someone who can choose jurisdiction. A health ministry comparing public systems, or a tour desk filling hotel beds, would build a different table. A country can have superb doctors and still sit mid-pack if a non-resident cannot enter, pay, schedule, or remain.

We did not rank on inbound patient volume. National definitions do not match: visits, patients, medical-purpose arrivals, same-day care. Volume is context only. We did not rank on a perception survey of people who have not used the system. We did not score prestige. We did not treat a CitizenX passport program as a proxy for hospital quality.

Comparative claims are written to be checkable. Swiss UWG Art. 3(1)(e) and the EU Unfair Commercial Practices Directive are the fairness standard: factual, relevant, not a pile-on.

Universe

The 21 countries are active medical-travel markets, jurisdictions CitizenX clients already use for residence or citizenship, or both.

Thailand, Türkiye, United Arab Emirates, Malaysia, Spain, Singapore, South Korea, Portugal, Costa Rica, Czechia, Mexico, Colombia, Taiwan, Greece, India, Germany, Panama, Japan, Israel, Switzerland, El Salvador.

Left out of the scored set on purpose:

• United States. It is the price comparator, not a tourism destination for most people who already live there. Next edition should add it as an unranked reference row.
• China, Saudi Arabia, Iran, Tunisia. Large or noisy inbound figures. Weak fit for the continuity and recourse test.
• Caribbean passport-program states, except as a warning later in this piece. Those instruments are passports. They are not tertiary hospital systems.

Pillars and weights

Score = 0.25Q + 0.20A + 0.15P + 0.15C + 0.15S + 0.10B

Quality is first because a cheap disaster is not a saving. Access is second because a superb hospital that will not take you next month is useless. Price cannot outvote quality. Continuity is the CitizenX pillar. Other indexes skip it. Follow-up care needs status. Safety is expected cost: street risk and malpractice. Breadth stops a cosmetic-only market from tying a city that can also do a bypass.

Rubric

Clinical quality starts with system-quality signals, then adjusts for the private international layer. Numbeo’s 2026 mid-year Health Care Index is the public comparable series (Taiwan 87.1, South Korea 83.0, Japan 79.9, Thailand 77.6, Spain 76.7, Mexico 72.3, Singapore 71.9, Türkiye 71.3, UAE 70.8, Malaysia 70.7, Switzerland 68.8, Colombia 68.9, India 65.6, Costa Rica 64.8, Greece 58.6). Newsweek/Statista World’s Best Hospitals lists are a presence check for academic hospitals, not a points table. JCI hospital-level density is a private-international process signal, not a quality ceiling. JCI measures process, not outcomes. Working 2026 hospital-level ranges: UAE 70–90, Thailand 60–65, India 40–47, Türkiye 38–42, Malaysia 12–16, Singapore 8–12, Colombia 11, Mexico about 9, Spain about 9, South Korea about 8, Israel about 5, Taiwan about 4. Confirm on the live directory before the next edition. Japan, Germany, Switzerland, and Taiwan often skip JCI because domestic regulation already binds their national champions. That is not a penalty. Anchor: 90+ is a system a paying foreigner can trust for complex work. 70–85 is a serious private market with variance. Below 60 is limited tertiary depth.

Non-resident access is entry plus a desk that will take you. Visa-free, e-visa, or a published medical visa a self-pay patient can get without a six-month consular process. Private hospitals with international desks, English-speaking coordinators, cash or card settlement. Scheduling in weeks, not public-queue months. OECD and Fraser Institute waits count against a country only where the foreigner would be stuck in that public line. Private wards in the same country can still score well. Anchor: 90+ means a competent patient can be on a ward within weeks on a tourist or e-visa. 60–75 is possible but frictional. Below 55 means the system is not built for this customer.

Price uses four procedures and mid-range published private-pay USD from 2024–2026 clinic and compiler ranges. Not a quote. Not an all-in medical package.

Illustrative destination bands. Full ranges sit in the workbook.

• India: hip ~$6,000–$10,000; CABG ~$8,000–$12,000; IVF ~$3,000–$5,000; implant ~$600–$1,500
• Türkiye: hip ~$8,000–$14,000; CABG ~$12,000–$20,000; IVF ~$2,000–$5,000; implant ~$700–$1,400
• Thailand: hip ~$12,000–$18,000; CABG ~$15,000–$25,000; IVF ~$3,000–$8,000; implant ~$1,200–$2,500
• Malaysia: hip ~$8,000–$12,000; CABG ~$8,000–$15,000
• Mexico: hip ~$10,000–$15,000; implant ~$750–$1,500
• Singapore, Switzerland, Israel, Japan: closer to Western private tariffs, so P is low by design

P is high when the basket is a large discount to the US cash band and the quality score is not in the basement. A $2,000 unknown clinic does not get a 100.

Continuity uses public pathways current at publication. Confirm the rule before anyone files. Direct citizenship or published long-stay and residence-by-investment routes that do not require a seven-year guess: Türkiye property-investment citizenship, El Salvador passport program, UAE long-stay residency, Portugal and Spain and Greece residence routes, Panama friendly-nations and pensionado, Malaysia MM2H, Thailand long-stay and LTR, Costa Rica and Colombia residence classes, Mexico temporary then permanent residence. Hard: Singapore, Switzerland, Germany, Japan, Taiwan, South Korea. Quality systems that do not sell status. India: OCI and medical visas are not treated as a Plan B passport. This pillar is not a ranking of passport programs. It asks whether healthcare and status can live in the same country.

Safety is personal security for the patient and a companion, plus a realistic malpractice or contract path after a bad outcome. Package-tour markets with weak follow-up lose points here on purpose. Geopolitical disruption that can cancel a treatment calendar also counts (Israel in this edition). Singapore, Japan, Switzerland, and Germany sit at the top. Mexico, Colombia, and Türkiye are discounted relative to their price scores. El Salvador is scored on current conditions, not a 2015 newspaper.

Specialty breadth asks whether the private market can do more than one product. High: cardiac, oncology, orthopedics, fertility, and diagnostics in the same city (Singapore, Germany, South Korea, Thailand, India, Switzerland). Mid: strong in two or three (Türkiye dental, hair, and cosmetic plus growing hospital groups; Spain fertility plus private surgery; Czechia fertility and dental). Low: thin tertiary menu (El Salvador, Panama).

Pillar scores, this edition

Integers are rubric judgments, not survey means. Change them in the workbook if a source moves.

Sources

• Numbeo, Health Care Index by Country, 2026 mid-year
• Joint Commission International accredited-organization directory; hospital-level counts cross-checked against 2026 compilations (Thailand Board of Investment, November 2025: 65 facilities)
• Newsweek / Statista, World’s Best Hospitals, latest published cycle, as a presence check
• Korea Health Industry Development Institute foreign-patient count for 2025 (2.01 million), via Yonhap / Korea Herald, 30 August 2026
• Türkiye official healthcare-visitor and revenue series for 2024 (1,506,442 visitors; about $3.02 billion) and 2025 volume notes
• Thailand foreign patient-visit estimates (Kasikorn / 2025 compilations around 3.07 million visits)
• India medical-purpose arrival counts (2024 about 644,000; 2025 507,244)
• Dubai Health Authority medical-tourist count for 2023 (691,478 in Dubai)
• Malaysia Healthcare Travel Council 2024 healthcare-traveller figure (1.6 million; RM2.72 billion)
• OECD Health at a Glance 2025 waiting-time tables (cataract, hip)
• Fraser Institute, Waiting Your Turn 2025 (Canada median 28.6 weeks referral to treatment)
• Published private-pay procedure ranges from clinic menus and 2025–2026 comparison tables. Ranges, not invoices
• Public government pages for residency and citizenship routes. Confirm before filing

What would move a rank next edition: a live JCI directory scrape by facility type; a named-hospital outcomes layer where a country publishes infection, revision, or 30-day mortality for private international patients; a United States cash-pay reference row; an insurance-versus-cash split for Singapore and Switzerland. A cleaner national inbound series still would not enter the weighted score until definitions converge.

What the order is doing

Thailand is first because the product already exists. Bumrungrad was the first hospital in Asia to take JCI accreditation, in 2002. Thailand’s Board of Investment put JCI-accredited hospitals and medical facilities at 65 as of November 2025. The private market is built for a foreigner who pays the bill. LTR and long-stay routes exist if recovery runs longer than ten days. Nobody needs to call it the best system on earth. It is good enough, and you can get in.

Türkiye is second for a reason volume lists skip. Hair and dental packages are the marketing. The index cares that you can enter and pay. If you want the jurisdiction rather than only the surgery, the residency and citizenship stack is already there. Price on the basket is aggressive. Safety and legal recourse are the discount. Read 80.6 as a trade.

The UAE is the continuity bet. Dubai alone reported 691,478 medical tourists in 2023. The federation also concentrates JCI hospital-level accreditations (working 2026 range: 70–90). You do not go to Dubai because bypass is cheap. You go because the hospital will take you, and a residency can outlive the discharge summary.

Singapore and Switzerland are expensive on purpose. Singapore General sits on Newsweek/Statista world lists. Swiss university hospitals do the same. Both score poorly on price and on how easy it is to remain. That is the right result. If the problem is a rare oncology protocol and money is not the constraint, this table should not pretend Bangkok is Zurich.

Taiwan has the highest clinical-quality subscore in the set. Numbeo’s 2026 mid-year Health Care Index puts Taiwan at 87.1, South Korea at 83.0, Japan at 79.9. Taiwan still finishes 13th. You still have to land, communicate, and stay. Quality with no access is a ranking for people who already live there.

India is the price column. Published private ranges for coronary bypass still cluster near $8,000–$12,000 against a US cash figure routinely quoted above $100,000. Apollo, Fortis, Medanta, and Narayana are real hospitals. OCI and medical visas are not a Plan B passport. If the only job is the operation, India belongs higher. That is not the only job.

El Salvador is last on purpose. CitizenX publishes an El Salvador passport program. The index does not pretend the passport is a hospital. Tertiary depth is thin. Score the jurisdiction for the service you need.

Volume is a different dataset

Inbound counts are useful. They are also not comparable.

Thailand counts foreign patient visits, including people who already live there. Malaysia reports "healthcare travellers." South Korea reports foreign patients: 2.01 million in 2025, of which 618,000 from China, 600,000 from Japan, 185,000 from Taiwan, and 173,000 from the United States, per KHIDI via Yonhap. India reports medical-purpose arrivals (about 644,000 in 2024, 507,244 in 2025). Tunisia’s 2.5 million figure for 2023 mixed hospital stays with same-day visits. Global Healthcare Accreditation has said for years that these series cannot be stacked into one league table.

Volume stays out of the weighted score. It is in the workbook as context.

Use the specialty, not only the headline

Cardiac and joints on a budget: India, then Malaysia and Türkiye.

Complex private care you will fly for and fly home from: Singapore, South Korea, Germany, Switzerland, Israel.

Fertility: Spain and Czechia belong in the same conversation as Thailand and Türkiye. Donor and testing law changes the destination more than the hotel.

Dental and hair: Türkiye and Mexico win on sticker price. Recourse after a bad implant is why they do not sit at 95. That cost lives in the 15% safety pillar.

Live in the country and use the system as a resident: Spain, Portugal, UAE, Malaysia, Costa Rica, Greece. Japan and Taiwan only after you have already solved status.

US cash prices are the outlier that makes the rest of the table possible. A single implant at $3,000–$6,000 in the United States against $700–$1,200 in Türkiye is not a rounding error. Neither is a Canadian median wait of 28.6 weeks from GP referral to treatment in the Fraser Institute 2025 survey. OECD 2024 median hip-replacement waits in several public systems run to many months. For a lot of households, private pay abroad is the alternative to a queue they did not vote for.

Limits

We do not run clinics. There is no CitizenX hospital list hiding under the table.

A score of 82.4 does not mean your surgeon in Bangkok is better than a named surgeon in Houston.

Several countries in the bottom half have excellent public hospitals for people who already hold the right ID. This ranking is not a verdict on those systems.

It is also not a visa-free table wearing a lab coat.

How to use it with a passport portfolio

Pick the job first.

Defined procedure, return ticket: quality, access, price. Thailand, Malaysia, Türkiye, South Korea, India.

"We might need this healthcare for a decade": quality, continuity, safety. UAE, Spain, Portugal. Singapore if you can obtain status. Costa Rica or Malaysia if the budget is finite.

A Plan B passport that happens to sit near decent private care is still a passport. Do not reverse that sentence. El Salvador, Vanuatu, and most Caribbean programs are travel and optionality instruments. Buy the hospital system separately. That is what this table is for.

CitizenX publishes all-in prices for passport programs on the site. Healthcare is separate, local, and yours. The Concierge can point you at the jurisdiction. A physician there still has to take you.