How to Predict Your Next Period (And How Far Off the Estimate Can Be)

Why the second half of the cycle is the stable one

A cycle has two phases. The first runs from the start of bleeding to ovulation and its length varies a great deal — stress, illness, travel and sleep all push ovulation later. The second, from ovulation to the next period, is much more constant, usually between 10 and 16 days. So when a cycle comes late, it is almost always because ovulation was late, not because the second half stretched.

This is why calendar prediction has a floor on its accuracy. A method that only counts days cannot see when ovulation actually happened, so it inherits every bit of variability from the first phase. Tracking basal temperature or ovulation tests narrows the window considerably, because they pin down the event the second half is measured from.

The 28-day cycle is an average, not a norm

Twenty-eight days is a convenient round figure that a minority of cycles actually match. Anything from 24 to 38 days is within the normal range, and using 28 for someone whose cycles run 33 days means a prediction that is wrong by five days every single month — enough to make an app look broken when the body is working perfectly.

Cycles also change across a lifetime. They are typically longer and more variable in the years after menarche, settle in the twenties and thirties, and shorten then become erratic in the run-up to menopause. An average calculated from cycles two years ago describes a body that no longer exists.

When a late period is worth investigating

One late cycle in an otherwise regular year is ordinary and usually needs nothing. What is worth a conversation is a pattern: cycles consistently shorter than 24 days or longer than 38, a gap of more than nine days between your shortest and longest cycle, bleeding that lasts more than eight days, or any bleeding between periods.

Take a record of your dates to the appointment. Six to twelve months of first days, written down, turns a vague impression into something a clinician can work with immediately, and it is the single most useful thing you can bring.

Frequently asked questions

Does the first day count as day one even if bleeding starts at night?

The usual convention is to count that day as day one if full bleeding starts before evening, and the following day if it starts late at night. A one-day difference will not change your average once you have six cycles, provided you apply the same rule every time.

Does hormonal contraception change the calculation?

Yes, and it usually makes it simpler. Combined pills produce a withdrawal bleed on a schedule set by the packet rather than a natural cycle, so the date follows the regimen. Other methods can make bleeding lighter, irregular or absent entirely, in which case day-counting no longer describes anything.

How accurate are period tracking apps?

As accurate as the regularity of the person using them, since most simply average past cycles the way you would by hand. They cannot detect a late ovulation, so their predicted date carries the same uncertainty as your own calculation. They should never be relied on as contraception.