The evidence that estrogen protects against coronary heart disease is biologically plausible, consistent, and strong. These benefits have not been established by a randomized trial, however, so that the degree of protection against heart disease might have been overestimated because estrogen users tend to be healthier than nonusers. A randomized trial to determine whether estrogen alone or in combination with progestin protects against coronary heart disease should be given a high priority. Progestins generally attenuate the effects of estrogen on the concentrations of HDLC. It is not known whether this effect also limits the beneficial effects of estrogen on the risk of coronary heart disease. Recent studies suggest that estrogen may protect against coronary heart disease in other ways besides favorably altering serum concentrations of lipoproteins and that progestins might not have adverse effects on the risk of heart disease. Currently, theoretical concerns that progestins might be harmful seem outweighed by the evidence that they protect against endometrial cancer in women who have a uterus. For these women, some may find the side effects of progestins to be so bothersome that they prefer to take estrogen alone. This approach is reasonable so long as the patient has periodic endometrial biopsies for early detection of pre-malignant or malignant endometrial changes. Women without a uterus should take estrogen alone. Women who take long-term estrogen therapy appear to have about a 30% greater chance of developing breast cancer. On the other hand, breast cancer that develops while taking estrogen therapy might have a slightly better prognosis. Quantitative comparisons of fatal conditions suggest that the benefits of long-term therapy outweigh the risks. But these comparisons assume that all causes of deaths are equally important and do not adequately take account of other psychologic and physical effects of hormone therapy.(ABSTRACT TRUNCATED AT 250 WORDS)