Objective: To evaluate the role of serum soluble endoglin in prediction of non-severe preeclampsia and severe preeclampsia in early second trimester (13-20 weeks) of pregnancy. Material and Methods: Prospective case–control study was carried out over a period of one and half years. Total 543 pregnant women were recruited but 43 were lost to follow up. Out of 500 pregnant women, 34 developed non severe preeclampsia (Group 2) and 16 developed severe preeclampsia (Group 3), and 51 healthy normotensive pregnant women (Group 1) were selected as controls. Serum soluble endoglin levels were measured by Human Eng (endoglin) ELISA kit. Endoglin levels were estimated by enzyme linked immune-sorbent assay technique. Results: Prevalence of hypertensive disorders of pregnancy (HDP) was 10.2%. The mean sEng levels were significantly higher in non-severe preeclampsia and severe preeclampsia as compared to controls (p<0.001). The mean sEng level was significantly elevated, almost more than doubled from baseline value, after the development of preeclampsia in both groups, non-severe preeclampsia (6.95 ± 0.19 vs. 13.65 ± 0.32, p<0.001) and severe preeclampsia (7.99 ± 0.27 vs. 17.81 ± 0.52, p<0.001) as compared to baseline values. At cut-off value of sEng >6.10 ng/ml not only discriminate non severe preeclampsia from controls with 82.35% sensitivity, 88.24% specificity, but also predict severe preeclampsia means severity of disease with 93.75% sensitivity, 88.24% specificity. To discrimination between non severe and severe PE, at cut–off value of >7.76 ng/ml, sEng had 68.75% sensitivity, 85.29% specificity, positive predictive value 68.7% and negative predictive value 85.3%. Conclusion: Thus sEng can be used as a potential biomarker to predict preeclampsia as early as in second trimester (13-20 weeks) of pregnancy.