BACKGROUND Septic shock is an important cause of death in pediatric intensive care units. Initial evaluations have shown that vasopressin may have a role in catecholamine refractory shock in adults. It is important to determine whether children with septic shock have deficiency of vasopressin. This will help in defining the role of vasopressin in septic shock. DESIGN Prospective cohort study. SETTING Pediatric intensive care unit of a tertiary care hospital in north India. PATIENTS Patients were children with septic shock, and controls were children with sepsis but no shock. STUDY DESIGN Vasopressin levels in plasma were determined by enzyme-linked immunosorbent assay for children with septic shock at diagnosis (baseline) and thereafter at 24, 48, and 96 hrs to determine the time trends. The baseline vasopressin values for children with septic shock were compared with those for children without shock. RESULTS The median (95% confidence interval) vasopressin level at baseline in children with septic shock was 116 (63.3-130.7) pg/mL, and in children with sepsis but no shock it was 106 (61.7-131.77) pg/mL. The median value for survivors was 76 (44.6-130.9) pg/mL, and for nonsurvivors, 118 (81.7-259) pg/mL (p = .16). The serial values also did not show any significant changes; the values at 24 hrs (n = 17), 48 hrs (n = 16), and 96 hrs (n = 15) were 105 (76.1-125.9), 105 (41.4-155.5), and 109.5 (54.9-154.8) pg/mL, respectively. CONCLUSIONS The results of our study suggest that vasopressin levels are elevated in children with septic shock and that serial values up to 96 hrs do not show any decline.