Objectives : The purpose of our study is to determine whether the fact that elevated level of maternal serum beta-hCG at second trimester could predict the occurrence of pregnancy-induced hypertension(PIH) is reasonable or not. Methods : From October 1997 to November 1998, 5427 women who had delivered over 20 gestational weeks at our hospital were reviewed. Among the fifty four hundred twenty seven women, three hundred sixty three pregnant women who had undergone second trimester triple marker screening test for fetal anomalies such as Down syndrome and neural tube defect were analyzed. The level of serum beta-hCG and outcomes of pregnancy were analyzed retrospectively and receiver operating characteristic curve (ROC curve) analysis was used to calculate the optimal cut-off value of beta-hCG. Results : The incidence rate of PIH was 6.1%, and the delivery time & neonatal birth weight, neonatal Apgar score at 1-minute and 5-minutes in PIH were significantly different compared with those in normal pregnancy ( 36.2±5.3weeks VS 39.2±1.3weeks, 2.7±0.8kg VS 3.3±1.3kg, 8.32±0.95 VS 8.94±0.31, 9.32±0.95 VS 9.94±0.31, P<.05). Mean value of beta-hCG of women with PIH was higher than the normal pregnant women (1.8±1.2 VS 1.1±0.7, p-value < 0.05), and pregnancies with beta-hCG above 2.0 MoM is at higher increased risk for PIH than those below 2.0 MoM (Odds ratio 7.73, 95% CI, 3.04-19.69). ROC curve implies that the level of hCG at 1.62 MOM is statistically important optimal cut-off level to predict PIH and that sensitivity, specificity, positive likelihood ratio and positive predictive value at 1.62 multiples of the median level are 54.5%, 85.9%, 3.87 and 20% respectively. Conclusion : Unexplained high level of beta-hCG in second trimester was associated with PIH, but positive predictive value of 20% at 1.62 MoM is unsuitable in screening PIH. So further large scaled and controlled prospective studies must be induced to determine whether unexplained high level of beta-hCG could predict PIH or not.