OBJECTIVE: Magnesium sulfate is used to prevent convulsions in preeclamptic women. It acts as a calcium antagonist but may also stimulate prostacyclin. Because magnesium sulfate readily crosses the placenta, we evaluated whether it might have a beneficial effect on placental blood flow. STUDY DESIGN: Isolated human placental cotyledons (n = 6) were perfused for 20-minute intervals with control Krebs-Ringer-bicarbonate buffer, 200 mu mol/L t-butyl hydroperoxide, magnesium sulfate (6 mEq/L), peroxide plus magnesium sulfate, and peroxide plus magnesium sulfate plus calcium chloride (6.25 mEq/L). Peroxide perfusion was used to stimulate thromboxane to induce vasoconstriction. Fetal perfusion pressure was continually monitored. Maternal and fetal effluent samples were analyzed for thromboxane and prostacyclin by their stable metabolites, thromboxane B-2 and 6-keto-prostaglandin F-1 alpha. RESULTS: Compared with control Krebs-Ringer-bicarbonate buffer perfusion, peroxide perfusion significantly increased (p < 0.05) vascular resistance (12.9 +/- 1.2 vs 21.1 +/- 2.6 mm Ho min/ml, mean +/- SE) and thromboxane B, secretion (fetal -0.22 +/- 0.08 vs 0.73 +/- 0.11 ng/min, maternal -1.5 +/- 0.4 vs 4.4 +/-: 0.7 ng/min). Subsequent perfusion with magnesium sulfate significantly attenuated (p < 0.05) peroxide-induced vasoconstriction (15.1 +/- 1.7 mm Hg min/ml), which was reversed by the addition of calcium (19.7 +/- 2.2 mm Hg . min/ml). Magnesium sulfate partially, but significantly (p < 0.05), inhibited the peroxide-induced increase in maternal thromboxane B-2 secretion (3.2 +/- 0.6 ng/min) but not fetal thromboxane B-2 secretion (1.1 +/- 0.3 ng/min). Magnesium sulfate did not affect 6-keto-prostaglandin F-1 alpha secretion. CONCLUSIONS: (1) Magnesium sulfate attenuates peroxide-induced vasoconstriction in the human placenta. (2) This effect is mediated by inhibition of thromboxane synthesis and antagonism of calcium.