Circulating levels of activin A are significantly increased in women with preeclampsia when compared with those with a normal pregnancy. The mechanisms underlying these increased levels are unknown. We undertook these studies to explore whether oxidative stress might be the mechanism. We exposed trophoblast explants, human umbilical vein endothelial cells (HUVECs) and peripheral blood monocytes to oxidative stress in vitro using xanthine/xanthine oxidase (X/XO), measuring activin A and isoprostane in conditioned media and mRNA for activin beta(A) in explants and HUVECs. We also measured isoprostane and activin A in serum from 21 women with preeclampsia and from 20 women with a normal pregnancy. Treatment with X/XO significantly increased 8-isoprostane production from placental explants, HUVECs and monocytes, indicative of oxidative stress, and significantly increased activin A output from placental explants (139.1 +/- 27.4 per mg wet weight vs 322.9 +/- 89.7 pg/ml per mg wet weight, P = 0.02) and from HUVECs (1.2 +/- 0.2 vs 3.2 +/- 1.8 ng/ml, P = 0.04). There was no effect on activin A output from monocytes. X/XO significantly increased beta(A) mRNA in placental explants but not in HUVECs. Maternal plasma levels of 8-isoprostane and activin A were significantly higher in women with preeclampsia when compared with controls (333.8 +/- 70 vs 176.3 +/- 26.2 pg/ml, P = 0.04 and 49.5 +/- 7 vs 13.1 +/- 1.2 ng/ml, P < 0.001 respectively). In the women with preeclampsia, but not in those with a normal pregnancy, circulating levels of 8-isoprostane and activin A were significantly and positively correlated (r(2) = 0.72; P < 0.001). These data suggest that oxidative stress may be one of the mechanisms underlying increased circulating activin A in preeclampsia.