How age arithmetic actually works

Calendar age is not a subtraction of two numbers, because months have unequal lengths. The standard method is a borrow algorithm, the same idea as long subtraction:

days = D2 − D1 ; months = M2 − M1 ; years = Y2 − Y1
If days is negative, add the number of days in the month before the target date and reduce months by one. If months is then negative, add 12 and reduce years by one.

That ordering matters. Borrowing days first, then months, gives the answer people expect; doing it in the other order produces off-by-one results near month boundaries.

A worked example

Date of birth 15 March 1990, measured on 28 August 2026:

ComponentCalculationResult
Days28 − 1513
Months8 − 35
Years2026 − 199036

No borrowing was needed, so the age is 36 years, 5 months and 13 days. The total-days figure is computed separately, because it cannot be derived from that triple:

SegmentDays
15 Mar 1990 to 15 Mar 2026 — 36 × 36513,140
Leap days in that span (1992, 1996, 2000, 2004, 2008, 2012, 2016, 2020, 2024)9
15 Mar 2026 to 28 Aug 2026166
Total lived13,315

Milestones that fall between birthdays

Because the total-days count runs independently of the calendar, several round numbers arrive on ordinary weekdays:

MilestoneApproximate age
1,000 days2 years 8 months
10,000 days27 years 4 months
1 billion seconds31 years 8 months
20,000 days54 years 9 months
30,000 days82 years 1 month

The leap-year rule in full

A year is a leap year if it is divisible by 4, except century years, which must also be divisible by 400. So 1900 was not a leap year and 2000 was. This is why any calculator that simply divides by 365.25 drifts: the true mean Gregorian year is 365.2425 days, and the 400-year cycle contains 97 leap days, not 100.

Why two calculators can give different answers

This is not sloppiness — the question is genuinely underdetermined at month ends. Take 31 January plus one month. Is the result 28 February, because February has no 31st and the day clamps to the last valid one? Or 3 March, because 31 days were added? Reputable date libraries disagree, and both behaviours are documented as correct within their own conventions.

The same ambiguity runs backwards. From 31 January to 1 March, is the elapsed time one month and one day, or 29 days? Counting whole months first gives the former; counting days gives the latter. A tool is only wrong if it is inconsistent with itself.

Two more sources of disagreement: whether the end date counts as a full day, and time zones. If your birth date is stored as midnight UTC and read in a zone behind UTC, the displayed date can shift by one day, which changes the answer by a day near midnight.

Where the calendar itself gets in the way

Dates before the Gregorian reform are ambiguous rather than merely old. Catholic Europe switched in October 1582; Britain and its colonies switched in September 1752, skipping eleven days so that 2 September was followed by 14 September; Russia switched in 1918 and Greece in 1923. A date recorded in a parish register from 1700 may be Julian, and no calculator can tell which without knowing the place. For genealogy, treat anything before the local switch as approximate.

There are also cultural systems that do not match this arithmetic at all. Korea historically counted a newborn as one year old and added a year at New Year rather than on the birthday, so a person could be two at a few days old. That reckoning was formally set aside in favour of international age in June 2023, but you will still meet both numbers in older documents.

Honest limits

Chronological age is a count of days, nothing more. It is not biological age, and it does not predict health, fitness or life expectancy for any individual — those depend on factors this tool has no access to. Any medical or fitness formula that takes age as an input, including target heart-rate ranges and BMI-for-age percentiles, is a population approximation and no substitute for a clinician.

Note too that some legal and financial contexts do not use the age you would calculate here. Several insurers price on age nearest birthday, which rounds up for anyone past the halfway point, so your quoted age can be one higher than your real one. Tax and pension thresholds sometimes use age at the end of a tax year instead.

Questions people actually ask

What happens if I was born on 29 February?

Mathematically your age increments on 1 March in non-leap years, because 29 February does not exist to be reached. Legally it varies by jurisdiction — some define the birthday as 28 February and some as 1 March, which occasionally matters for a document that must be signed at exactly 18.

Why does the day count not equal years times 365?

Leap days. Over 36 years there are typically nine of them, so the total is nine days higher than the naive product. Over a century the gap is about 24 days.

Can I find the weekday I was born on?

Yes, and it is deterministic — the Gregorian calendar repeats on a 400-year cycle, so the weekday for any date follows from the date alone. It has no bearing on anything beyond curiosity.

Is my date of birth stored anywhere?

No. The calculation runs in your browser and the date you enter is never transmitted or saved.

Keep exploring Gen Code Tools

Every tool comes with a written guide, and every category is one click away.