Tamoxifen protects against myocardial infarction through mechanisms that are poorly understood. We studied the effects of tamoxifen and another antiestrogen, toremifene, on the production of vasoconstrictive endothelin-1 and of vasodilatory nitric oxide in 44 postmenopausal patients with breast cancer. These started treatment, in randomized order, with either tamoxifen (20 mg/day; n = 25) or toremifene (40 mg/day; n = 19). Plasma samples collected before treatment and after 6 and 12 months of both regimens were assayed for endothelin-1 with a specific radioimmunoassay and for nitrite/nitrate with a method based on the Griess reaction. The antiestrogen group as a whole showed a fall in endothelin-1 at 6 months (5.9 +/- 3.3%; p = 0.06) (mean +/- SE) and at 12 months (7.1 +/- 5.5%; p = 0.03). This fall was solely due to toremifene, the use of which was associated with falls in endothelin-1 at 6 months (12.9 +/- 4.7%; p = 0.01) and 12 months (9.2 +/- 6.2%; p = 0.06). The antiestrogen regimen failed to affect plasma nitric oxide significantly but nevertheless the ratio between nitric oxide and endothelin-1 rose by 31.6 +/- 13.3% at 6 months and by 35.6 +/- 15.3% at 12 months in the antiestrogen users, an effect similar in the tamoxifen and toremifene groups. We conclude that antiestrogens may protect against myocardial infarction by preventing the release of endothelin-1 and by shifting the balance between nitric oxide and endothelin-1 to the dominance of the former. Our data predict that toremifene and tamoxifen at the doses studied here will provide similar cardiovascular protection.