Women who develop preeclampsia during pregnancy excrete less calcium than healthy pregnant women. Whether this reduction in calcium excretion precedes or follows hypertension is unknown. We prospectively measured urinary calcium excretion in 103 consecutive nulliparous women at risk for preeclampsia and presenting for prenatal care before 24 weeks' gestation. Serial 24-hour urine specimens were obtained at 10-24 weeks, 25-32 weeks, and 33 weeks to term. After delivery, the charts were reviewed for the presence of preeclampsia and gestational hypertension. At the first collection, patients who later developed preeclampsia excreted less urinary calcium (169 ± 30 mg/24 hours; mean ± standard error of the mean) than those who remained normotensive (298 ± 15 mg/24 hours) (P<.05); this reduction persisted throughout gestation. Using a receiver operator curve, we calculated a predictive threshold calcium value for hypertension of 195 mg/24 hours. The difference in the incidence of preeclampsia between pregnant women with calcium excretion values at or below 195 mg/24 hours (87%, 95% confidence interval 52-98%) and those with values above that level (2%, confidence interval 0.3-8%) was highly significant (Fisher exact test,P<.0001). The 95% lower limit of relative risk for preeclampsia in patients with a calcium excretion equal to or below 195 mg/24 hours in the first collection was 9.4. These observations suggest a pathophysiologic role for altered urinary calcium excretion in women with preeclampsia that may contribute to the early identification of patients at risk for this disease.