The effects of four medical treatments have been assessed on menstrual blood loss (MBL) and endometrial prostaglandin (PG) concentrations in 30 women with objectively confirmed menorrhagia. Patients were randomly treated with danazol, 200 mg daily (n = 6), mefenamic acid, 500 mg three times daily during menses (n = 8), norethisterone, 5 mg twice daily from day 15-25 of the cycle (n = 8) or a progesterone-impregnated coil releasing 65 micrograms progesterone daily (n = 8). Endometrial biopsies were obtained in the mid-luteal phase before and after treatment in 23 cases, and assayed for PG content using radioimmunoassay. Treatment with norethisterone had no effect on either MBL or the concentration of PGs in the endometrium. MBL was significantly reduced after treatment with mefenamic acid (P = 0.05, n = 6) and the progesterone coil (P less than 0.05, n = 6), and was reduced in each of 4 cases treated with danazol in whom endometrial biopsies were available. Although there was no consistent change in endometrial PG concentrations in either the mefenamic acid or danazol groups, the lower MBL after insertion of the progesterone coil was associated with a reduced endometrial content of PGE, PGF2 alpha and "total" PG (6oxo PGF1 alpha + PGE + PGF2 alpha)-P = 0.05. Whereas the cyclooxygenase inhibitor mefenamic acid is likely to exert its effect on endometrial PGs at the time of menstruation itself, the continuous administration of progesterone throughout the menstrual cycle could result in both an impairment in estrogen receptor generation leading to reduced estrogen-mediated cyclooxygenase activity, and an increase in endometrial PG metabolism.