How birthday rarity is calculated.

Where do the birth records come from?

The main U.S. result combines two daily-count datasets compiled by FiveThirtyEight: CDC/NCHS records and Social Security Administration records.

1. Use CDC/NCHS counts for 1994–1999.
2. Use SSA counts for 2000–2014.
3. Combine the resulting 7,670 calendar dates. The overlapping 2000–2003 years are counted once.

These are daily birth totals, not names or individual birth certificates. The upstream datasets are attributed under CC BY 4.0. Administrative coverage differs between the sources.

Explore the U.S. birthday calendar, or choose another available country and time period. Each view identifies its own source; separate countries are never combined into a worldwide ranking.

What does “rarer than” mean?

The main result adds up recorded births for each month and day. “Rarer than” compares that total with the other observed birthday dates: it is the percentage of other dates with strictly more recorded births. Tied dates do not count as more common.

“More common than” makes the reverse comparison. A top-percent badge includes the whole tied group before awarding a tier. A singular “rarest” or “most common” label requires a unique result.

What does “1 in” mean?

Divide the total births in the selected sample by the births on your month and day, then round to a whole number. For September 9, that is 85,712,738 divided by 259,228: approximately 1 in 331 births.

This answers a different question from the percentile. Birth odds describe the share of recorded births; the percentile compares birthday dates.

Why is February 29 so rare?

February 29 is the rarest birthday by total births in the main U.S. sample. It occurs only five times in this 21-year window. Every other month and day occurs 21 times.

The main rarity result uses actual totals, so those fewer opportunities count. In Explore, “Births per occurrence” divides by the number of times a date occurred. That lets you compare an average leap day with an average occurrence of another birthday.

Both measures are useful, but they answer different questions. An unobserved February 29 in a single-year view is missing data, not a day with zero births.

Is this a worldwide birthday ranking?

No. The main result describes historical U.S. birth records. It is not a count of everyone alive today, a current-year estimate, or a worldwide ranking.

September 9 is the most common birthday in this U.S. sample. The order can change when you choose another country, period or measure. Partial-year records do not provide a balanced comparison of all birthdays, so whole-year rarity results are withheld for those selections.

Can birthdays reveal a conception date?

A backward date estimate is an illustration, not a person's actual conception date. The app's optional model shifts and smooths birth counts; it cannot establish individual histories or the cause of a seasonal pattern.

What is the rarest birthday without leap day?

Christmas Day, December 25, has the fewest recorded births among annual dates in this U.S. sample: 138,628, or about 162 per 100,000 births. September 9 has 259,228, about 1.87 times as many. The top-ten lists offer a “Without leap day” option.

What month has the most birthdays?

In the main U.S. dataset, August has the most total births (7,624,626), while September has the highest average per day (about 11,811). A month with 31 days gets more opportunities to accumulate births. The birth-month chart lets you compare both measures.

Why do September and the holidays stand out?

Birth timing reflects several influences, including seasonality and scheduled deliveries. The UK Office for National Statistics discusses late-September peaks and holiday dips in its own England and Wales data. Its explanation is context, not proof of any individual conception date or the cause of a pattern in this U.S. sample.

Does the birthday paradox mean someone will share my birthday?

Those are different probabilities. For 23 independently sampled people, the observed U.S. birthday distribution gives about a 50.8% chance of any matching pair. The chance that at least one of those 23 people matches a specified September 9 birthday is about 6.7%. The app calculates both using the same recorded distribution; the group excludes you. Cambridge’s NRICH explains the distinction behind the birthday paradox.

Questions about the data?

Contact @adriannalakatos with your question or data request.