What Is a 100-Year Flood? Return Periods and AEP Explained

“1-in-100-year flood” is probably the most misunderstood phrase in flood engineering. It does not mean a flood that arrives on a hundred-year schedule, and it does not mean that after one occurs you are safe for another century. It is a statement of probability — and once you read it that way, some surprising numbers fall out. A property with a 100-year flood exposure has roughly a 40% chance of experiencing that flood at least once over a 50-year period.

This article unpacks the definition properly, shows the small amount of maths behind it, and explains why engineers increasingly prefer to talk about annual exceedance probability instead of return periods.

The definition: annual exceedance probability

A 1-in-100-year flood is a flood magnitude with a 1% chance of being equalled or exceeded in any single year. That 1% is the annual exceedance probability, or AEP. The relationship is simply:

AEP = 1 / T

where T is the return period in years. So:

Return period	AEP (chance in any year)
1:10 year	10%
1:20 year	5%
1:50 year	2%
1:100 year	1%
1:200 year	0.5%

Each year is an independent draw. The river does not remember last season. Two 1:100-year floods can occur in consecutive years — or in the same year — and the statistics are not violated, merely unlucky.

The maths that matters: risk over a design life

The annual probability understates how the risk feels over the life of a development, because exposure accumulates year after year. The probability of experiencing at least one exceedance over n years is:

R = 1 − (1 − 1/T)ⁿ

Applying this to a 1:100-year flood:

Exposure period	Chance of at least one 1:100-year flood
1 year	1%
10 years	9.6%
30 years	26%
50 years	39.5%
70 years	50.5%
100 years	63.4%

Two results here are worth pausing on. First, over a typical 50-year building design life, the “100-year” flood is closer to a coin flip weighted 40/60 than to a remote possibility. Second, even over a full 100 years, the chance is 63% — not 100%. The name “100-year flood” guarantees nothing about any particular century.

You can run these numbers for any return period and exposure window with our return period and flood risk calculator.

Where the 1:100-year magnitude comes from

The flood magnitude assigned to a given return period is an estimate, not a measurement. For a specific river it is derived from statistical analysis of recorded flows where a suitable gauging record exists, and from rainfall-based rainfall–runoff methods — such as the Rational Method or the SCS-SA method — where it does not, which is most South African sites.

That means the 1:100-year flood for your site can change as records lengthen, methods improve, or the catchment itself changes through development. It is a best current estimate with uncertainty attached — one of the reasons regulators favour conservative design events and why a defensible methodology matters as much as the number itself.

Why this framing matters in practice

• Planning and floodlines. South African municipalities most commonly regulate against the 1:100-year event for building plans and township layouts, with the 1:50-year line sometimes shown alongside it. Our post on 1:50 vs 1:100 year floodlines covers which applies where.
• Communicating risk to non-engineers. “A 40% chance over the life of your building” lands very differently from “the 100-year flood” — and it is the same fact. For buyers, lenders, and insurers, the design-life framing is usually the honest one.
• Higher-consequence infrastructure. Because risk compounds over time, long-lived and high-consequence structures — bridges, dams, major culverts — are checked against rarer events (1:200-year and beyond, up to the Regional Maximum Flood or PMF) than ordinary developments.

Common misconceptions, quickly

• “We just had the 100-year flood, so we’re safe for decades.” No — the AEP resets to 1% next year, every year.
• “Outside the 1:100-year floodline means no flood risk.” It means roughly less than 1% annual probability under current estimates. Larger floods exist, and estimates carry uncertainty.
• “The 1:100-year flood is a fixed physical quantity.” It is a statistical estimate that can and does get revised.

If a floodline or flood probability is shaping a decision on your site — a purchase, a layout, a building plan — a site-specific flood risk assessment turns these generic probabilities into numbers for your actual property.