Preeclampsia, a cardiovascular disorder of late pregnancy, is characterized as a low-renin hypertensive state relative to normotensive pregnancy. Because other nonpregnant low-renin hypertensive disorders often exhibit and are occasionally dependent on elevated arginine vasopressin (AVP) secretion, we hypothesized a possible use for plasma AVP measurements in the prediction of preeclampsia. Copeptin is an inert prosegment of AVP that is secreted in a 1:1 molar ratio and exhibits a substantially longer biological half-life compared with AVP, rendering it a clinically useful biomarker of AVP secretion. Copeptin was measured throughout pregnancy in maternal plasma from preeclamptic and control women. Maternal plasma copeptin was significantly higher throughout preeclamptic pregnancies versus control pregnancies. While controlling for clinically significant confounders (age, body mass index, chronic essential hypertension, twin gestation, diabetes mellitus, and history of preeclampsia) using multivariate regression, the association of higher copeptin concentration and the development of preeclampsia remained significant. Receiver operating characteristic analyses reveal that as early as the sixth week of gestation, elevated maternal plasma copeptin concentration is a highly significant predictor of preeclampsia throughout pregnancy. Finally, chronic infusion of AVP during pregnancy (24 ng per hour) is sufficient to phenocopy preeclampsia in C57BL/6J mice, causing pregnancy-specific hypertension, renal glomerular endotheliosis, proteinuria, and intrauterine growth restriction. These data implicate AVP release as a novel predictive biomarker for preeclampsia very early in pregnancy, identify chronic AVP infusion as a novel and clinically relevant model of preeclampsia in mice, and are consistent with a potential causative role for AVP in preeclampsia in humans.