Short answer

Most period trackers predict your next period by averaging the length of your recent cycles and counting forward from the first day of your last period. Predictions improve as you log more cycles, but they stay estimates, because cycle length changes from month to month and the day you ovulate varies even more.

Key facts

  • In 612,613 cycles logged in a fertility app, the average cycle lasted 29.3 days, and a typical user's cycle length varied by about 2.6 days (one standard deviation) from cycle to cycle.[1]
  • In the same data, the days before ovulation (the follicular phase) averaged 16.9 days, so ovulation often came later than the textbook day 14.[1]
  • In a study of 949 women, 34% believed they had a 28-day cycle, but only 15% did.[2]
  • Cycle apps in that study predicted the actual day of ovulation no more than 21% of the time.[2]

What information does a period tracker use?

The core input is the date each period starts. Your cycle length is the number of days from the first day of one period to the first day of the next.[3] Some apps also use how long you bleed, plus optional signals such as ovulation (LH) tests and basal body temperature (BBT) that say more about when you ovulate.

SignalWhat it tells the appLimits
Period start datesYour usual cycle length and how much it variesSays nothing directly about ovulation
LH urine testsA rise in luteinizing hormone, which comes shortly before ovulation[4]You need to test daily around the expected window
Basal body temperatureTemperature stays slightly higher after ovulation, so a lasting rise suggests ovulation has happened[5]Confirms ovulation only afterwards; broken sleep or illness can blur the pattern
Symptoms and moodPersonal patterns, such as cramps or low mood before a periodUseful context, but not a reliable ovulation marker

How is the next period date calculated?

  1. The app collects the lengths of your recent cycles.
  2. It averages them to estimate your typical cycle length. Some apps give recent cycles more weight.
  3. It adds that length to the first day of your last period to estimate when the next one starts.
  4. To estimate the fertile window, it counts back from the predicted period by the luteal phase (the days after ovulation, which averaged 12.4 days in the large app study[1]) to estimate ovulation, then marks that day and the five days before it.[6]

If your cycles vary a lot, a single predicted date hides that uncertainty. Showing a range of likely days is a more honest way to present the same estimate.

How accurate are period predictions?

For the next period, accuracy depends mostly on how regular your cycles are. If your cycles usually land within a day or two of each other, an average-based prediction will usually be close. If they swing by a week or more, it will often miss.

Predicting ovulation is harder. In a prospective study of 221 women, the fertile window fell entirely within cycle days 10 to 17 for only about 30% of them, and on every day from day 6 to day 21 at least 10% of women were in their fertile window.[7] The researchers concluded that the timing of the fertile window can be highly unpredictable, even if their cycles are usually regular.[7]

That is why ovulation estimates based on dates alone are weak: a 2018 study found that ovulation day varies a lot for any given cycle length.[2]

Why do irregular cycles get a date range?

With PCOS, and during perimenopause, cycle length can change a lot from one month to the next.[8][9] A range shows that uncertainty instead of hiding it, and makes a "late period" alert less alarming when your cycle is simply variable.

Can you use period predictions as birth control?

No. A predicted fertile window is an estimate, and ovulation can come earlier or later than predicted.[7] Preeo is not a method of birth control and must not be used to prevent pregnancy. If you want to avoid pregnancy, talk with a clinician about contraception.