Women who wish to become pregnant are usually advised to have intercourse during a theoretic “fertile window” of the five days before ovulation and the day of ovulation. These guidelines are not reliable because the day of ovulation may vary from cycle to cycle in an individual woman in addition to variations among women. Wilcox and colleagues studied women who were planning pregnancy to establish new guidelines for the fertile window in healthy women.

The authors used a study of early pregnancy to investigate ovulation in more than 200 healthy women. The volunteers were 25 to 35 years of age, and most were white and college-educated. Women with fertility problems were excluded from the study. The data collected at enrollment included information about the length and regularity of the women's menstrual cycles. Once birth control was discontinued, the women collected the first urine sample of the day for hormone analysis and kept records of the days when intercourse and menstruation occurred. The day of ovulation was calculated by a change in the ratio of metabolites of estrogen and progesterone in the urine samples. Each menstrual cycle was dated by the first day of menstrual bleeding.

Overall, 136 of the 221 women conceived during the study. Ovulation occurred as early as the eighth day and as late as the 60th day of the menstrual cycle. On days 12 and 13, more than one half of the women were in the fertile period, but 17 percent were already in the window by the seventh day of their cycle. The 16 percent of women who reported irregular periods tended to ovulate later and at more variable times. Nevertheless, even women who reported regular cycles had a probability of up to 6 percent of being in the fertile window on the day that menstruation was expected to start.

The authors conclude that current assumptions about fertile periods are unreliable. On each day between days 6 and 21, there is at least a 10 percent probability of being in the fertile window, and sporadic late ovulation cannot be predicted. Current clinical guidelines are valid for only about 30 percent of healthy women. Timing intercourse to achieve or avoid pregnancy cannot be done with precision.