Extrahepatic portal vein obstruction (EHPVO) affects 20–30% of all patients with portal hypertension in India. The etiopathogenesis of this disease is unknown. In retrospective studies, umbilical vein cannulation and sepsis have been alleged to cause portal thrombosis. This prospective study was undertaken to detect clinically and by serial sonography whether thrombosis and consequent obstruction of the splenoportal venous system develops after umbilical vein catheterization and sepsis. Forty-seven children who had undergone exchange transfusion for hyperbilirubinemia, belonging to two different age groups, were studied. Twenty-two new-borns (Group A) were studied within 4 weeks of birth and later at 3-month intervals until the age of 12 months and subsequently at 24 months. Another group of 25 children (Group B), 1− to 5-year-olds who had earlier undergone exchange transfusion and 15 healthy newborns (Group C) were also screened. In Group A, seven (29%) neonates had septicemia, and in five, the splenic vein could not be initially visualized. The splenic vein was not seen in five of the 15 newborns in Group A who had no umbilical sepsis. Twenty-one of the 22 neonates in Group A were folio wed-up, and the splenic vein was well visualized and found patent on subsequent ultrasound in all of them. In Group B, four (16%) children had umbilical sepsis. None of these or the other Group B children showed any thrombosis or obstruction of the splenoportal system. The portal vein and its branches and the splenic vein were visualized in all healthy (Group C) neonates. In conclusion, our results demonstrate that umbilical vein catheterization and sepsis do not lead to development of portal vein thrombosis, at least, in the hospital.