I claimed the exclusion for three years and then my accountant filed me on the foreign tax credit because it saved money. Have I revoked?

Almost certainly yes, if you had foreign earned income in that year and you qualified under either the bona fide residence test or the physical presence test. Filing a return that omits the exclusion while an election is live is treated as turning the election off, whether or not anyone intended that result and whether or not a revocation statement was attached. The saving in that year is real but it purchased a multi-year lock-out on re-electing. Your first step is to establish the exact year in which the omission occurred, because the lock-out window runs from that year and you need to know when it expires.

Does a year in which I did not qualify for the exclusion count as a revocation?

No. Failing the qualification tests is not an election decision, so a year in which you were neither a bona fide resident of the UK nor met the physical presence requirement does not revoke anything. The difficulty is evidential rather than legal. If your file contains no day count, no residence analysis, and no contemporaneous note explaining why the exclusion was absent, you are asking the IRS to accept a characterisation years after the fact with nothing behind it. Where a non-qualifying year is likely, document the reason at the time it happens rather than when it is challenged.

Can I get permission to re-elect early, and how hard is it?

Permission requires the consent of the Commissioner, obtained through the private letter ruling process, with the procedure and user fee published on IRS.gov. It is a formal submission, not a statement attached to a return, and the fee is usually smaller than the professional cost of preparing it properly. Consent is discretionary and the rule exists specifically to stop taxpayers alternating between the exclusion and the credit to take the better outcome each year. A request grounded in a genuine change of circumstances reads very differently from one whose real explanation is that the credit produced a smaller number that year.

My UK effective rate is well above the US rate every year. Why would I ever keep the exclusion alive?

Because you are buying an option on your own future, and the years in which the exclusion matters are the years your current rate comparison cannot see. A move to a Gulf or Singapore posting, a sabbatical, a period of overseas workday relief, or a wind-down to part-time work all produce years in which your foreign effective rate falls below the US rate and the credit stops covering you. If any of those are plausible within the lock-out horizon, claiming a partial exclusion in a year where the credit is marginally better preserves the election at modest cost. The exclusion and the credit can coexist within a single year on different slices of income.

Does the housing exclusion revoke separately from the earned income exclusion?

Yes. They are distinct elections with distinct revocation consequences, which means you can be locked out of one while the other remains live, and you can revoke one without touching the other. This matters for clients with high London housing costs, where the housing element can carry real value independently of the earned income exclusion. It also means a preparer who quietly drops the housing claim has made a decision with its own multi-year tail. Your file should record the status of each election separately, for each spouse.

We file jointly. If our return drops the exclusion, are we both revoked?

Each spouse holds their own section 911 election, so the answer depends on how the return was actually prepared and whose income was excluded. It is entirely possible for one spouse to be revoked while the other's election remains intact, and this is common where one spouse earns substantially more than the other. The lower-earning spouse is often the one for whom the exclusion has the most future value, because their income is more likely to fall near or below the annual exclusion amount in a given year. Review the election status separately for each of you rather than treating the joint return as a single position.

FEIE Revocation Trap: The Five-Year Lock Explained

The foreign earned income exclusion under section 911 is not a switch you may flip on and off each filing season according to whichever method produces the lower US liability that year. It is an election, and elections under section 911 carry a revocation consequence: once you revoke, you are barred from re-electing for a period of tax years running from the year of revocation, unless the Internal Revenue Service grants you permission to re-elect sooner. That permission is not obtained by ticking a box or attaching a statement to a return. It is obtained through a private letter ruling, a formal process with a user fee, a submission burden, and no guarantee of a favourable answer. The trap for Americans living in Britain is that revocation frequently happens by accident. A taxpayer who has claimed the exclusion and then simply files a subsequent year on the foreign tax credit method, without excluding any foreign earned income, has in substance revoked the election for that year and every year that follows within the lock-out window. Because UK effective tax rates on employment income usually sit comfortably above US rates, the credit method looks costless in most years, and it usually is. The damage surfaces later, in the one year the client most needed the exclusion: a low-income year, a sabbatical, a secondment to a lower-tax jurisdiction, a year dominated by UK-exempt or lightly taxed compensation. That is the year the exclusion would have been worth the most, and that is precisely the year it is unavailable. The resolution is prophylactic rather than curative: treat the section 911 election as a multi-year commitment, decide deliberately whether to keep it alive, and never allow a return to drop the exclusion without a documented reason for doing so.

Why do advisers keep treating the exclusion as an annual choice?

The habit is understandable. Most cross-border tax software models the exclusion and the foreign tax credit as competing computations, presents the resulting numbers side by side, and invites the preparer to select the better outcome. The modelling exercise is genuinely useful. What the modelling exercise omits is the temporal dimension. A comparison that spans a single tax year cannot see the cost of a revocation, because that cost is deferred and contingent. It only materialises in a future year whose facts are not yet on the preparer's desk.

There is also a structural reason. For the large majority of Americans in the United Kingdom, the credit method is the better answer in ordinary years. UK income tax and National Insurance on employment income generate creditable foreign taxes at rates that exceed the effective US rate on the same income, and the excess flows into carryover. The exclusion, by contrast, is capped at the annual exclusion amount, sterilises the underlying foreign tax so that it cannot be credited against the excluded portion, and interacts awkwardly with the stacking rule that taxes non-excluded income at the rate that would apply if the excluded income were still in the base. Run the comparison in a normal year for a well-paid client in London and the credit wins. Run it again the next year and it wins again. The preparer, quite reasonably, keeps choosing the winner. Nobody records that the first of those choices was an act of revocation with a multi-year tail.

The final reason is that the exclusion is popularly described as something you claim rather than something you elect. IRS.gov sets out the mechanics of making the election and the consequences of revoking it, and the language is clear enough on its face, but the framing in practice is transactional. A form is completed, a number is excluded, a return is filed. The election vocabulary sits in the background until the day someone needs it back.

What actually counts as a revocation?

There are two routes to revocation, and only one of them looks like a decision. The first is an affirmative revocation: the taxpayer attaches a statement to a filed return declaring that the section 911 election is revoked. This is deliberate, documented, and rare. The second is functional. A taxpayer who has an election in place, who continues to have foreign earned income, and who files a return that does not claim the exclusion has effectively turned the election off. Whether one calls this a revocation, a failure to continue the election, or a deemed revocation is a matter of vocabulary. The practical consequence is the same: the taxpayer who later wants the exclusion back cannot simply resume claiming it.

The subtlety worth understanding is that the election, once made, is designed to persist. It carries forward from year to year without any affirmative renewal. That persistence is a convenience while the taxpayer wants the exclusion and a hazard the moment a return is prepared on a different basis. The absence of the exclusion on a return is not a neutral act. It is a change of position with respect to a live election.

Two adjacent situations are often confused with revocation and should be distinguished. A year in which the taxpayer has no foreign earned income at all does not present the question, because there is nothing to exclude and no position being taken. A year in which the taxpayer fails the physical presence test and the bona fide residence test is different again: the exclusion is unavailable as a matter of qualification rather than election, and the taxpayer has not chosen anything. The difficulty is that these situations are easy to assert after the fact and hard to substantiate. If your file does not show which of them applied, the Service is not obliged to accept the characterisation you prefer years later.

How does the lock-out period actually operate?

The lock-out runs by reference to the tax year in which the revocation took effect, and it covers a run of subsequent tax years during which the taxpayer may not make a new section 911 election for the same category of exclusion. The exclusion for foreign earned income and the exclusion or deduction for foreign housing are treated as separate elections, which means a taxpayer can be locked out of one while retaining the other, and can revoke one without automatically revoking the other. That separation is a planning point and a trap in equal measure, because a preparer who drops the housing element while retaining the earned income exclusion has made a decision with its own multi-year consequence.

Re-election within the lock-out window is possible only with the consent of the Commissioner. Consent is sought through the private letter ruling process. The IRS publishes the ruling procedure and the applicable user fee schedule on IRS.gov, and the fee is not the principal cost. The principal cost is the professional time required to build a submission that explains the facts, the reason for the earlier revocation, and why the request is not an attempt to whipsaw the fisc by alternating between methods to harvest the better of the two in each year. That last point is the whole purpose of the rule. The lock-out exists to prevent exactly the year-by-year optimisation that the modelling spreadsheet invites.

Rulings are granted. They are not granted casually, and the fact patterns that succeed tend to involve a genuine change in circumstances rather than a change in arithmetic. A taxpayer who revoked because a corporate reorganisation moved them onto a different payroll structure is in a different position from one who revoked because the credit produced a smaller number that year. Practitioners should assume that a submission whose honest answer to "why did you revoke?" is "because it saved tax that year" will not be well received.

Which UK fact patterns turn the lock into a real loss?

The sabbatical or unpaid leave year. A client on partial pay for part of a year may have foreign earned income well below the annual exclusion amount and correspondingly little UK tax to credit, particularly once the UK personal allowance and the progressive rate structure absorb most of the earnings. The exclusion would have removed the income from the US base entirely. Without it, the client pays US tax on income that carries almost no creditable foreign tax.
The transitional year on a move to a lower-tax posting. Americans in Britain who accept a regional role covering the Gulf, Singapore, or Switzerland frequently find their effective foreign rate falls below the US rate. The credit method, which was ample cover in the UK, becomes partial cover overnight. The exclusion would have been the primary shield in exactly those years, and the carryover generated during the high-tax UK years may be the wrong basket to help.
The year dominated by income that the UK does not tax or taxes lightly. Certain remittance-basis positions, overseas workday relief arrangements, and periods of non-residence can leave a client with US-taxable foreign earned income that has attracted little or no UK charge. HMRC guidance on GOV.UK sets out when UK employment income falls outside the UK charge, and each of those situations is a year in which the foreign tax credit has nothing to work with.
The self-employed consultant with a lean year. A US citizen contracting through a UK personal service company or as a sole trader may draw modest earned income in a year of business investment. Self-employment tax remains due regardless of the exclusion, but the income tax exposure is materially different depending on whether the exclusion is live.
The early-retirement or wind-down year. A client reducing to part-time work before ceasing UK employment altogether may spend two or three years in the band where the exclusion would have removed the entire earned income figure from the US base. This is a common shape and an easy one to see coming, which makes it a poor one to be surprised by.
The trailing spouse with independent earnings. Each spouse holds their own section 911 election. A joint return that drops the exclusion can revoke for one spouse, both, or neither depending on how the return was prepared, and the analysis for a lower-earning spouse frequently differs from the analysis for the primary earner.

How should the election be governed rather than merely computed?

The correct posture is to treat the section 911 election as a standing position in the client file, on the same footing as a treaty election, an entity classification, or a mark-to-market election. Standing positions are reviewed deliberately, changed deliberately, and documented when changed. They are not re-decided every year by whichever preparer has the file.

That means the annual comparison should produce two outputs, not one. The first is the current-year number. The second is an explicit statement of what happens to the election. If the recommendation is to file on the credit method for a client with a live election, the memorandum should say that this revokes the election, identify the lock-out window, and record that the client accepted the consequence. In practice, once that sentence appears in writing, a meaningful proportion of clients decide differently.

It also means that the exclusion should sometimes be claimed in a year when it is not optimal. Preserving the election has an option value. If the client's forward view includes any of the fact patterns above within the lock-out horizon, claiming a modest exclusion in a year where the credit would have been marginally better is cheap insurance. The exclusion and the credit are not mutually exclusive within a year: income above the exclusion amount can still carry creditable foreign tax, subject to the allocation and stacking rules. A partial claim keeps the election alive.

What should be in the file before you change methods?

A dated record of the year in which the section 911 election was first made, for each spouse separately, and a note of whether the housing element was elected alongside it. Reconstructing this from old returns years later is expensive and sometimes impossible where prior-year workpapers have been discarded.
A forward projection covering at least the length of the lock-out window, identifying any year in which foreign earned income is expected to fall near or below the annual exclusion amount, or in which the expected foreign effective rate falls below the US rate. This does not need precision. It needs to identify shape.
A written client instruction, signed or confirmed by email, acknowledging that filing without the exclusion revokes the election, that re-election requires IRS consent obtained by private letter ruling, and that consent is discretionary. Advisers who omit this step and are later asked to explain a lost exclusion have no defence worth the name.
A note distinguishing years of non-qualification from years of non-election. Where a client fails both the bona fide residence and physical presence tests, the file should show the day count or residence analysis that produced the failure. This is the single most useful document if the characterisation of a past year is ever contested.
An assessment of whether the housing exclusion or deduction should be treated separately from the earned income exclusion, given that they are distinct elections with distinct revocation consequences and distinct value depending on the client's London or regional housing cost.

What if the revocation has already happened?

Begin by testing whether it happened at all. Examine each intervening year and ask whether there was foreign earned income, whether the client qualified, and whether the exclusion was omitted as a positive choice or was simply unavailable. A year with no qualification is not a revocation, and a taxpayer whose exclusion lapsed because they returned to the United States for a period and failed both tests may be in a better position than the returns suggest.

If the omission was a genuine error rather than a considered position, an amended return for an open year that claims the exclusion may repair the position, because the election is then continuous rather than interrupted. The window for this is finite and depends on the limitation period for the year in question, so the diagnosis needs to happen early. This is one of the reasons a change of adviser should always trigger a review of standing elections rather than only the current year's numbers.

Where the revocation is real and the lock-out is live, the analysis becomes a cost comparison. Model the US liability across the remaining locked years without the exclusion. Compare that figure to the cost of a private letter ruling request, including the user fee published on IRS.gov and the professional time to prepare it, discounted by an honest probability of success given the reason for the original revocation. For many clients the exposure will not justify a ruling and the answer is to absorb the cost and diarise the year the lock expires. For a client facing several low-tax years inside the window, the arithmetic can point the other way. Either way, the decision should be made on the numbers rather than on the discomfort of having been caught by a rule that most of the market still describes as an annual choice.