A normal cytologic pattern of a vaginal smear following oral administration of 2.5 mg of conjugated estrogens three times daily for 5 days during early pregnancy was more reliable in prognosticating a good pregnancy outcome than a vaginal smear unrelated to the taking of estrogen. But an abnormal vaginal cytologic pattern following such administration of estrogen was no more accurate in predicting abortion than an abnormal smear obtained prior to the use of estrogen.