Flood Return Periods & Annual Probability Explained

A flood return period is the average length of time between floods of a given size at a given place. That sounds like a timetable, and it is not one. A "1-in-100-year flood" does not arrive once a century: it means a flood of that size has a 1% chance of happening in any single year, every year, whether or not one happened last winter. Hydrologists also call it a recurrence interval, and the Environment Agency increasingly writes it the other way round, as an annual exceedance probability. Getting the number straight matters, because it sits behind flood zones, planning requirements, insurance premiums and the risk attached to the property you are buying or building on. It is also only half the picture: the return period describes the flood, while a property's own annual probability of flooding describes the building. If a search has flagged flood risk on a purchase, an independent flood risk survey when buying a house turns that headline figure into what it actually means at the threshold.

A 1-in-100-year flood has a 1% chance of occurring in any given year, and a 26% chance of striking at least once during a 30-year mortgage.

The calculator below converts between the two ways of writing the same number, and compounds the annual chance over any period you choose.

What is a flood return period?

A flood return period is the average interval between floods that reach or exceed a particular size at a particular place. A river might have a 1-in-100-year flood level, a 1-in-1,000-year flood level, and so on, each larger and rarer than the last. The figure comes from fitting a statistical distribution to gauged river flow and rainfall extremes, so it estimates long-run flood event frequency rather than measuring a fixed cycle.

Three terms describe the same quantity, and a flood risk assessment may use all of them in the same chapter without contradiction.

• Return period. The common British usage, and the phrasing behind "1 in 100 years". The glossary entry for a flood return period sets out the definition in one line.
• Recurrence interval. The same average spacing, sometimes written as an annual recurrence interval. It is more common in academic hydrology and in North American practice, and it means nothing different.
• Annual exceedance probability. The same figure inverted into a yearly percentage, and the form the Environment Agency now prefers. The glossary covers annual exceedance probability alongside it.

The useful move is to read the return period as its inverse. If a flood recurs on average once every 100 years, then in any single year the chance of it happening is one in a hundred, a 1% annual probability. That inverse is the number that actually governs how flood risk is mapped and managed in England and Wales, from the flood zone a site falls in to the standard a scheme has to be designed against.

Return period and annual probability are the same fact stated two ways: the annual chance is one divided by the return period.

What does a 1-in-100-year flood actually mean?

It means a 1% chance every year, not a flood that turns up on a hundred-year schedule. The phrase is one of the most misread terms in flood risk, because "1-in-100-year" sounds like a calendar promise when it is really a dice roll repeated annually.

Picture a bag holding one hundred numbered balls, with ball number one standing for the flood. Each year one ball is drawn and put back. Drawing the one does not empty the bag or change next year's draw, so the odds are exactly the same the following January. A flood can occur two years running, or twice in one winter, and the long-run average still holds. The "100 years" is the average spacing over a very long record. It is not a countdown between events.

Flood return period and AEP calculator

The calculator converts a return period into its annual exceedance probability and back, then compounds that annual chance over a period of your choosing. The arithmetic is set out in full in the section below it, and the working assumptions are on the reference tab.

How to calculate a return period: the AEP formula, worked both ways

The conversion is a single division. Annual exceedance probability is the reciprocal of the return period, so AEP = 1 ÷ T, and the return period is T = 1 ÷ AEP. Working an example both ways takes seconds, and the calculator above does it alongside the compounded odds.

1. Take a return period, say T = 200 years.
2. Divide one by it: 1 ÷ 200 = 0.005.
3. Multiply by 100 for the percentage: a 0.5% AEP.
4. To reverse it, divide 100 by the AEP percentage: 100 ÷ 0.5 = a 1-in-200-year return period.

AEP is the Environment Agency's preferred formal term, and the word "exceedance" is the precise part: it is the chance of a flood meeting or exceeding a given size in any year, not hitting it exactly. A 1-in-100-year flood has a 1% AEP; a 1-in-200-year flood a 0.5% AEP. The percentage form is now standard in technical flood risk assessments and Environment Agency documents, precisely because it heads off the "once a century" misunderstanding.

One step sits behind the arithmetic and is easy to miss. The return period itself is not measured; it is estimated, by fitting a statistical distribution to the annual maximum flows in a gauged record, and the methods for doing so are set out in the Flood Estimation Handbook while the underlying records come from the National River Flow Archive. A 1-in-1,000-year estimate drawn from a forty-year record is an extrapolation, and it carries wide confidence limits. That is one reason a site-specific figure usually comes from hydraulic modelling of the reach rather than off a map.

1-in-100-year flood = 1% annual chance = 0.01 AEP; 1-in-30 = 3.3%; 1-in-50 = 2%; 1-in-1,000 = 0.1%.

Return period vs a property's annual probability of flooding

These are two different questions, and confusing them is where most of the trouble starts. A return period is flood-centric: it describes the size and rarity of an event on a river or a stretch of coast. A property's annual probability of flooding is property-centric: it is the chance that water actually reaches the building in a given year, which depends on ground level, distance from the watercourse, the flood source and any defences. A house can sit beside a river with a 1-in-100-year flood level and still have a low annual probability of flooding, because the floor sits well above that level.

The Environment Agency's online checker reflects exactly this split. Its long-term flood risk service gives a headline rating for an area rather than a surveyed figure for a threshold, and the Flood Map for Planning shows the event-based flood zones. Neither tells you the depth of water at the door. They also answer different questions, a distinction unpacked in the guide to the long term flood risk map and the Flood Map for Planning.

This is the distinction a screening result cannot make on its own. Moving from an area rating to a building's real annual probability of flooding needs levels taken on site: that is what an independent flood risk survey for a property purchase provides, and what a flood risk assessment for planning establishes for a development. Where the modelled depth matters as well as the probability, the Environment Agency now publishes banded flood depth data for rivers and sea alongside the extents.

Why the odds don't reset after a flood

Probability is not prediction, and a flood does not "use up" the risk for the years that follow. If a 1-in-100-year flood hits an area this year, the chance next year is still 1%. There are no ninety-nine safe years in the bank. This is where the timetable misreading does real harm, lulling people into thinking a recent flood buys long-term safety.

• It does not mean one flood per century. The interval is a long-run average, not a fixed gap between events.
• It does not reset the clock. Each year is an independent draw at the same odds, whatever happened last year.
• It is not a ceiling. A bigger, rarer flood than the "1-in-100" can still occur, and often has.
• It is not the defended risk. The return period describes the flood, not whatever protection sits between it and the door.

Clusters of rare events are the clearest illustration. Two 1-in-100-year floods in consecutive years are improbable, not contradictory. They are now studied in their own right, and the piece on back-to-back flooding sets out why. Boscastle in 2004 makes the same point from the other direction: a catchment small enough that an extreme rainfall event produced a flood far beyond anything in its record, as the account of the Boscastle flood explains.

If a home sits in a flood zone despite never having flooded, this is usually why. The map describes probability, not history. The explainer on why a house can be in a flood zone if it has never flooded unpacks that gap in full.

Return period and annual probability: the conversion scale

Every return period converts to a percentage chance by dividing one by the number of years. The table below shows the common bands used in UK flood risk work, each with the plain-English reading that goes with it.

How likely is a 1-in-100-year flood over the life of a mortgage?

More likely than the "1-in-100" label suggests, because small annual chances compound over time. A 1% chance in any one year is low. Stack thirty of those years together and the odds of seeing at least one such flood climb steeply.

The arithmetic is the complement of the flood not happening, P = 1 − (1 − AEP)ⁿ, where n is the number of years. Over a 30-year mortgage, a 1% annual chance compounds to a 26% probability of at least one 1-in-100-year flood, better than one in four. Over ten years the figure is about 10%. Over fifty it approaches 40%. Planning practice guidance takes the lifetime of a residential development as at least 100 years, and over that span the same 1% annual chance reaches 63%.

Small yearly odds add up: a 1% annual chance becomes a 26% chance over 30 years, and 63% over the 100-year lifetime planning guidance assumes for housing.

The annual number stays reassuringly small; the lifetime number is the one a buyer, lender or developer should actually weigh. It is why flood risk increasingly shapes mortgage lending decisions, and why it shows up in the evidence on whether flood risk affects house prices.

How flood zones are built on return periods

England's Flood Map for Planning divides land into flood zones defined directly by annual probability. It is the same return-period arithmetic turned into planning categories. Since 17 August 2026 the definitions have sat in Annex F, table 1 of the National Planning Policy Framework itself, alongside the new Chapter 18 flood policies F1 to F9; before that they lived only in planning practice guidance. The thresholds did not change. Where a report should cite them did, and that is one of several shifts covered in the review of the August 2026 NPPF changes to flood risk and drainage.

• Flood Zone 1 (low). Less than a 0.1% annual probability of river or sea flooding, which is rarer than a 1-in-1,000-year event.
• Flood Zone 2 (medium). Between 1% and 0.1% a year from rivers (1-in-100 to 1-in-1,000), or between 0.5% and 0.1% from the sea.
• Flood Zone 3a (high). 1% or greater a year from rivers (1-in-100 or more frequent), or 0.5% or greater from the sea (1-in-200 or more frequent).
• Flood Zone 3b (functional floodplain). Land where water flows or is stored in a flood, normally at a 3.3% annual probability or greater, which is a 1-in-30-year event, or land designed to flood. The boundary is agreed locally with the Environment Agency, and the functional floodplain is not shown separately on the map.

That 3.3% figure is the one most often misquoted, usually as 5%, so it is worth taking from the source. Annex F is careful to add that the identification of functional floodplain should take account of local circumstances and not be defined solely on rigid probability parameters, which means a local strategic flood risk assessment can draw the boundary differently for its own catchments in agreement with the Environment Agency, and that latitude is where a good deal of the confusion over the figure begins.

Two catches trip people up. The zones deliberately ignore flood defences, so a defended property can still sit in Zone 3 on the map. A scheme of new defences does not move the line, as still in a flood zone after new flood defences explains. And the zones carry no climate change uplift of their own. The guide to flood zones 1, 2 and 3 in England covers how the bands are applied in practice, and why flood maps and real-world risk don't always match covers the limits of the mapping. Where a site sits in Zone 2 or 3, the zone is also what triggers the Sequential and Exception Tests under policies F5 and F6.

Wales runs a different framework. Technical Advice Note 15, updated in March 2026, sets its own flood zones and requires a flood consequences assessment rather than an English-style FRA, though the underlying probabilities are read the same way.

Why climate change is making rare floods more frequent

Historic return periods are backward-looking. They are calculated from past records, so they carry no allowance of their own for a wetter, more volatile climate. A flood the record still calls "1-in-100" may already be arriving more often as rainfall patterns shift, and the UK Climate Projections behind the allowances point the same way, and the Annex F note is explicit that the mapped zones account for neither defences nor the possible effects of climate change.

That uplift is handled separately. For planning and design, the Environment Agency publishes climate change allowances: percentage increases to peak river flow and peak rainfall, varying by management catchment and time horizon, applied on top of the historic figure. The practical upshot is that today's 1-in-100-year flood level is not tomorrow's. The guides to climate change allowances for planning and how increasing rainfall is changing flood risk in the UK go further, sea level rise does the same job for the tidal case, and the residual flood risk explainer covers the risk that always remains once defences and design have done their work.

A historic 1-in-100-year flood level carries no climate uplift on its own; that allowance is added separately, and it is why rare floods are becoming less rare.

What return periods mean when you're buying, building or insuring

For most people the return period only matters through a decision: a purchase, a planning application, or an insurance renewal. It means something slightly different in each.

• Buying a home. The return period behind a "high risk" flag gives the odds, not the certainty, and a property-level survey shows how a flood of that size would affect the building rather than the postcode.
• Building or developing. The design flood return period sets the standard a scheme must be assessed and protected against. That is 1% for rivers and surface water, 0.5% for the sea, plus the climate change allowance, and it feeds straight into the flood risk assessment the application needs, scoped to the Environment Agency's guidance on flood risk assessments for planning.
• Insuring. A higher annual probability generally means higher premiums or excesses, and property-level evidence can correct a screening result that overstates the risk.

Because the map speaks in probability, the number on its own rarely settles the question. A postcode search on the flood risk map is a starting point, and the drainage and flood risk calculators cover the arithmetic that goes with it, but neither is a site assessment. Where a return period or flood zone is affecting a purchase, a planning application or a policy, Unda's chartered consultants can establish the real annual probability of flooding at a specific site, from an independent flood risk survey when buying a house to a flood risk assessment for planning or an assessment for insurance. Start a quote and a response follows within the hour.

Frequently asked questions