Progestin is often added to regimens of estrogen therapy in postmenopausal women to reduce the risk of endometrial hyperstimulation, but it may cause undesirable metabolic effects. Therefore, a low dosage is recommended. At present, the only way to determine whether the dosage of progestin is causing the desired secretory transformation of the endometrium is by endometrial sampling, which is invasive. We studied 102 postmenopausal women undergoing estrogen therapy who also took a progestin for 12 days each month, and we correlated the day of onset of bleeding with the endometrial histology over a three-month period. A bleeding pattern suitable for interpretation was observed in 96 women. Regardless of the preparation and dosage of the estrogen and progestin used, wholly or predominantly proliferative endometrium was always associated with bleeding on or before day 10 after the addition of progestin; wholly or predominantly secretory endometrium, or a lack of endometrial tissue, was associated with bleeding on day 11 or later. We conclude that the bleeding pattern reflected the histologic condition of the endometrium and that adjustment of the dosage of progestin so that regular bleeding is induced on or after day 11 may obviate the need for endometrial biopsy.