Floods

What Is a 100-Year Flood? The 1% Chance Explained

By the Water Level Atlas team · Last reviewed Oct 21, 2026

A "100-year flood" is a flood so large that it has a 1 percent chance of being equaled or exceeded in any given year. It does not mean it happens once every 100 years. Because the chance resets every year, it can happen two years in a row, and over 30 years the chance of at least one is about 26 percent.

What does "100-year flood" really mean?

Hydrologists describe floods by their annual exceedance probability (AEP): the chance that a flood of a given size will be equaled or exceeded in any one year. The flood with a 1 percent AEP is commonly called the "100-year flood". USGS notes the name was an attempt to simplify the idea, but it often leads people to think such a flood comes once a century. Talking about the chance in a given year is clearer.

The "100 years" is a long-term average, called the recurrence interval. If you could watch a river for thousands of years with nothing changing, floods of this size would average out to about one per century. In any real record, the gaps between them can be much shorter or much longer.

Why can it happen two years in a row?

Each year starts with the same 1 percent chance. A big flood last year does not use up the chance or make this year safer. USGS puts it simply: a rare flood does not reduce the chances of another rare flood within a short time.

If each year is treated as independent, the chance of 1-percent floods in two particular back-to-back years is 1 percent × 1 percent = 0.01 percent. Small, but not zero.

What is the chance over 30 years?

The usual question is about a 30-year mortgage. The answer is about 26 percent. USGS and FEMA both use this figure; FEMA describes its Zone A flood areas as having a 1 percent annual chance of flooding and a 26 percent chance over the life of a 30-year mortgage.

Here is how it is worked out. The chance of no 1-percent flood in one year is 99 percent (0.99). The chance of none in 30 independent years is 0.99 multiplied by itself 30 times, which is about 0.74. So the chance of at least one is:

1 − 0.99³⁰ ≈ 1 − 0.74 = 0.26, or about 26 percent.

The same formula gives other periods (rounded):

Even over a full 100 years, it is not certain. These numbers assume the true chance stays at 1 percent each year, which is the textbook assumption.

How is the 100-year flood estimated?

USGS looks at all the annual peak flows recorded at a gauge, one per water year, and fits them to a statistical model to estimate the chance of floods of different sizes. See annual peak flow and record crests for how the peak series is built.

USGS says ten or more years of data are needed for this kind of analysis, and that more years are better. With a short record, the estimate is much less certain.

Why does the 100-year flood level change?

Because it is an estimate from past data, it changes as the record grows. USGS says plainly that as more data come in, the level of the 100-year flood will change.

USGS also lists other reasons the estimate can shift over time:

• new dams and other river impoundments;
• more paved and roofed (impervious) surfaces in the watershed;
• other land-use changes;
• long-term climate patterns.

USGS also marks peak records that were affected by regulation, diversion, urbanization or channel changes. In short, yesterday's "100-year flood" may not be today's.

Does this tell me if my home will flood?

Not on its own. Flood insurance maps and flood risk decisions are made by FEMA, local floodplain managers and other officials, using their own studies. For live conditions, follow the National Weather Service and local authorities.

Using this on Water Level Atlas

We do not calculate our own 100-year flood values. River pages show the gauge's highest annual peak from USGS and the historic crests and flood stages NOAA publishes, so you can see how today's level compares with past floods at that spot. Flood status always comes from NOAA, with a link to the official page; see flood stages explained and our methodology page.