There is evidence that aspirin in low doses favorably influences the course of pregnancy-induced hypertension, but the mechanism, although assumed to involve suppression of the production of thromboxane by platelets, has not been established. We performed a randomized study of the effect of the long-term daily administration of 60 mg of aspirin (n = 17) or placebo (n = 16) on platelet thromboxane A2 and vascular prostacyclin in women at risk for pregnancy-induced hypertension. Low doses of aspirin were associated with a longer pregnancy and increased weight of newborns. Serum levels of thromboxane B2, a stable product of thromboxane A2, were almost completely (greater than 90 percent) inhibited by low doses of aspirin. The urinary excretion of immunoreactive thromboxane B2 was significantly reduced without changes in the level of 6-keto-prostaglandin F1 alpha, a product of prostacyclin. Mass spectrometric analysis showed that aspirin reduced the excretion of the 2,3-dinor-thromboxane B2 metabolite--mainly of platelet origin--by 81 percent and of thromboxane B2, probably chiefly of renal origin, by 59 percent. The urinary excretion of 6-keto-prostaglandin F1 alpha and of its metabolite 2,3-dinor-6-keto-prostaglandin F1 alpha was not affected. Low doses of aspirin only partially (63 percent) reduced neonatal serum thromboxane B2. No hemorrhagic complications were observed in the newborns. Thus, in women at risk for pregnancy-induced hypertension, low doses of aspirin selectively suppressed maternal platelet thromboxane B2 while sparing vascular prostacyclin, but only partially suppressed neonatal platelet thromboxane B2, allowing hemostatic competence in the fetus and newborn.