Birthday paradox calculator

Enter a group size to see the chance that two people share a birthday, using how often each date actually occurs in U.S. births.

How many people are in the room?

Real U.S. birth frequencies, 2000–2014, leap day included. Everything is calculated on this page.

In a group of 23

50.8%

chance that at least two share a birthday

365 equal days
50.7%
Pairs to compare
253
No shared birthday
49.2%

An even chance at

23 people

50.8% with real birth data

The 365-day formula gives 50.7%

Sharing your birthday

226+ people

for an even chance someone shares September 12

December 25: 447+ · February 29: 1,030+

Almost certain at

57 people

99.0% chance of a shared birthday

70 people: 99.9%

The birthday paradox, with real birthdays

How the chance grows with the group

Real birthdays are not spread evenly, which nudges every group’s chance up slightly.

0%50%100%11023405780
People in the group (horizontal) and the chance of a shared birthday (vertical). Purple marks 23, the first group size above an even chance.

Chance that at least two people share a birthday.

PeopleReal U.S. birth data365 equal daysPairs of people
52.7%2.7%10
1011.7%11.7%45
1525.3%25.3%105
2041.2%41.1%190
2350.8%50.7%253
2556.9%56.9%300
3070.7%70.6%435
4089.2%89.1%780
5097.1%97.0%1,225
5799.0%99.0%1,596
6099.4%99.4%1,770
7099.9%99.9%2,415

Each date’s share of 62,187,024 U.S. births (SSA via FiveThirtyEight, CC BY 4.0). The chance that everyone differs is calculated exactly from those shares; the 365-day column is the textbook formula.

Why 23 people is enough

The chance that no two people share a birthday shrinks with every pair, and pairs multiply quickly: 23 people make 253 pairs. With real U.S. birth frequencies, a group of 23 has a 50.8% chance of at least one shared birthday; at 57 people it is 99.0%, and at 70 it is 99.9%.

Real birthdays are uneven

Textbook answers assume 365 equally likely days. Real births are not spread evenly: more fall in late summer, fewer on holidays, and February 29 occurs only once every four years. Uneven dates make matches slightly more likely, so each group’s chance is a little higher than the textbook figure, 50.8% against 50.7% at 23. This calculator uses each date’s share of 62,187,024 U.S. births from 2000 to 2014.

Sharing your own birthday

Matching one particular date is far harder. You would need about 253 other people for an even chance of someone sharing your birthday if all days were equal. With real data it takes 226 people for September 12, the most common birthday, 447 for December 25 and 1,030 for February 29. See how common your own date is on the birthday heatmap.

How it is calculated

For each group size, the chance that everyone’s birthday differs is calculated exactly from the 366 date shares; one minus that is the chance of a shared birthday. Nothing you enter leaves the page. Counts come from SSA records compiled by FiveThirtyEight (CC BY 4.0).

Good questions

What is the birthday paradox?

It is the surprise that a small group is likely to contain two people with the same birthday: with 23 people the chance passes 50%. It feels wrong because we think of our own birthday, but 23 people form 253 pairs, and any pair can match.

How many people do you need for a 50% chance of a shared birthday?

23. With real U.S. birth frequencies the chance is 50.8%; the textbook 365-day formula gives 50.7%. With 22 people it is just under half.

What is the birthday problem?

The same question stated as a probability problem: how many people are needed before a shared birthday becomes likely? This calculator answers it for any group from 2 to 100 people.

What are the odds someone shares my birthday?

Much lower, because only your date counts: you need about 253 other people for an even chance with 365 equal days. With real data it takes 226 for September 12, the most common birthday, 447 for December 25 and 1,030 for February 29.