When not to buy international health insurance.

In short: International health insurance is the wrong purchase in five situations: a compulsory national scheme already covers you and holding a private policy does not exempt you from it; a social-security treaty already covers you; the local residency rule rejects international policies by construction; you are permanently settled inside one country’s system; or the premium is larger than the risk you are actually carrying.

We advise on international health insurance for a living, so read this article knowing what it costs us to write: there are five situations in which the honest answer to “which international policy should I buy?” is none. Not as a provocation — as the direct consequence of rules we have verified against primary sources across twenty-three countries. The product is genuinely excellent at what it is for. Here is when it is not for you.

1. Does a compulsory system already have me?

The Netherlands fines you (€529.74, twice, then signs you up itself — the CAK’s own figures), Switzerland gives you ninety days, Germany enrols employees through payroll, Japan requires application within fourteen days and back-bills two years. In all four, the international policy you proudly hold does not exempt you — liability follows your situation, not your existing cover. Buying international cover on top means paying twice for the base layer.

Follow one household through it, because the shape is always the same. A couple lands in Rotterdam in March holding a family international plan bought before the move, and reasonably assume it is the answer. It is not: Dutch basic insurance is compulsory from the moment they become liable. The CAK’s ladder then runs its course — a fine of €529.74, a second one three months later, and finally the CAK arranging cover itself and withholding the standard premium, €172.70 a month, from their income for twelve months. All the while the international premium keeps leaving the account. At the end of that year they have paid an international premium they chose, a Dutch premium they did not, and two fines — and the international policy prevented none of it, because it was never the thing the law was asking for. The Dutch clock itself — who must insure, and by when — is in the Dutch four-month clock.

The intelligent spend in compulsory countries is small and local: the Dutch supplementary for dental, the Swiss supplementary bought young, the German Zusatzversicherung, the narrow Japanese top-up against the 30% share. The base is not yours to choose; stop paying for a second one.

2. Does a treaty already cover me?

The least-known money-saver in our practice: social-security agreements. A Swiss employee seconded to Tokyo can remain in Swiss insurance and be excluded from Japanese NHI with a certificate of coverage — the Japan–Switzerland agreement has been in force since 2012, verified against the official guides. A Swiss national moving to Montreal may skip Quebec’s three-month RAMQ wait entirely under the Québec–Switzerland agreement. UK, Irish and eight other passports carry reciprocal Medicare into Australia. Every one of these is cover you already own through your nationality or secondment status, and every year people buy twelve months of international insurance to duplicate it. Before pricing any policy, spend thirty minutes on the treaty question: does my passport or my posting already do this?

3. Is the local requirement the wrong shape for it?

Spain rejects international policies for the NLV almost by construction — the sin copagos test fails any policy with an excess, at any size, and demands an insurer authorised in Spain. Costa Rica’s residency requires the CAJA specifically; no private policy substitutes, by statute. In both countries an international policy bought “for the visa” fails the visa. What compliance actually requires is a local product — and once you hold the compliant local product, the international layer must justify itself against a much smaller remaining gap, which it often cannot.

4. Am I permanently in one country, inside its system?

International insurance’s defining feature is portability — the same cover following you across borders. A person settled for good in one country, enrolled in its public system, is paying for portability they will never use. The French resident with a carte vitale needs a mutuelle, not a global policy; the settled Portuguese resident needs a small private layer against SNS waits, not worldwide cover. The test is honest self-description: if the realistic scenario is this country, indefinitely, buy the local supplementary product built for exactly the gaps this system leaves. Keep international cover only while the “indefinitely” is genuinely uncertain — which, granted, it often is in year one, and often is not in year five. Resize when the answer firms.

5. Would the money be better held than spent?

For the young, healthy, childless arrival in a country with a functional public system and modest private prices, a maximal international policy can cost more per year than the realistic worst season of self-paying. We would never say this of the United States, of evacuation-shaped geographies, or of anyone with dependants or conditions — there the tail risk is the whole point. But a 28-year-old inside the German, French or Portuguese system, buying a premium global plan “to be safe”, is often insuring a risk the public system already carries, with money that would serve them better as an emergency fund. Insurance is for the losses you cannot carry. Price what you are actually carrying first.

What does not buying cost later?

The counterweight, stated as plainly as the argument above, because an adviser who gives you only one side of this has given you half a reading.

Health insurance is one of the few purchases where when you first bought matters as much as what. Cover is underwritten against the person applying on the day, so a continuous policy held since your thirties carries the terms of a thirty-something’s health history — and that person stops existing the moment there is a gap. Re-applying at forty-five means answering the questions again, with the intervening decade in the answers: the back that was investigated, the medication started, the scan that found something incidental. What comes back can be an exclusion for that body part, a loading on the premium, or a decline. Waiting periods generally restart too, which matters most for maternity, where the wait is long enough to remove the benefit from the year you actually need it.

None of that argues for buying cover you do not need. It argues for distinguishing two decisions that get run together: do not buy and cancel what you hold are different recommendations. If you are choosing not to start, the cost is future underwriting risk you can weigh honestly. If you are cancelling a well-built policy held for years, you are also discarding an underwriting position you cannot repurchase — and that deserves more thought than the premium line alone suggests. In our review work, “keep the old policy, drop the duplicate layer” is a far more common conclusion than “cancel everything”.

When is the product clearly right?

None of this argues against cover — it argues against duplicate cover, non-compliant cover, and portability without a border in sight. The international product earns its premium brilliantly where its features are real: multi-country lives, exclusionary systems like Singapore’s, evacuation geographies, application gaps, and the USA question — which is its own category, since the SIP Health Cost Index 2025 ranks the United States first of fifty countries at about $17,969 a year for comparable cover, the only row in the index priced to include American treatment. Where treatment costs that much, “hold the money instead” stops being a strategy.

The discipline is only ever to name the gap before buying the thing that fills it. Naming the gap is most of what a review is — and it is why the most common outcome of ours is a smaller policy than the person arrived with.