This report defines the nature of the molecules responsible for the increased plasma plasminogen activator inhibitor (PAI) activity in preeclamptic patients and the relationship of these inhibitors to the severity of placental damage in preeclampsia. Clinical groups consisting of pregnant women with either severe preeclampsia or chronic hypertension with superimposed severe preeclampsia, as well as normal pregnant and nonpregnant women, were analyzed in a panel of functional and immunologic assays for PAI-1 and PAI-2. Pure severe preeclamptic patients in their third trimester showed a significant increase in both antigenic (136 ng/mL) and functional (5.76 U/mL) type 1 PAI (PAI-1) as compared with normal third-trimester pregnant women (34.8 ng/mL and 2.57 U/mL, respectively). In contrast, antigenic (186 ng/mL) and functional (5.76 U/mL) levels of type 2 PAI (PAI-2) were significantly lower in the pure severe preeclampsia group as compared with the values of the normal pregnant group (269 ng/mL and 9.58 U/mL, respectively). The patients with chronic hypertension and superimposed severe preeclampsia exhibited PAI-2 levels comparable to those of the pure preeclamptic group, whereas their antigenic and functional PAI-1 levels were intermediate (94 ng/mL and 3.25 U/mL, respectively) between the normal pregnant and the pure preeclamptic groups. During early puerperium of both normal pregnant women and patients, plasma PAI-1 antigen and activity decreased within one day to approximately the levels detected in normal nonpregnant women, while PAI-2 levels remained elevated for over 11 days. Similar results were obtained in plasma samples obtained from citrated blood and blood collected with an anticoagulant/antiplatelet mixture, suggesting that increased PAI-1 levels in preeclamptic patients were not due to platelet activation in vitro. In preeclamptic patients, a positive correlation between birth weight and PAI-2 values was observed (r = .64, P less than .05), whereas birth weight was inversely correlated with both PAI-1 levels and total PAI activity (r = -.6, P less than .005 and r = -.76, P less than .005 respectively). Preeclamptic patients with extensive placental infarction exhibited higher plasma PAI activity (24.1 U/mL v 11.6 U/mL) and PAI-1 values (305 ng/mL v 80.9 ng/mL) than preeclamptic patients without extensive placental infarction. In contrast, PAI-2 levels were reduced in preeclamptic patients with infarction in comparison with those of patients without infarction (141 ng/mL v 212.9 ng/mL). Our data indicate that increases in the level of PAI-1 accounts for the high plasma PAI activity in severe preeclampsia as measured using single-chain t-PA.