Measurement of the severity of surgery would greatly facilitate the design and interpretation of studies in neonates undergoing surgery. A scoring method, based on the amount of blood loss, superficial dissection, and visceral trauma, the site and duration of surgery, cardiac surgical factors, and associated stress factors for surgical neonates, was formulated and applied to 94 neonates undergoing surgery. Perioperative management was standardized for all patients and hormonal-metabolic variables were measured in blood samples drawn preoperatively at the end of the operation, and at six, 12, and 24 hours after operation. The stress scores were correlated significantly with the plasma epinephrine (P less than .0001), norepinephrine (P less than .0001), insulin (P less than .001), glucagon (P less than .005), and cortisol (P less than .02) responses, and with changes in blood glucose (P less than .0001), lactate (P less than .0001), pyruvate (P less than .0001), and alanine (P less than .005) during and after operation. Discriminant function analysis was used for further validation and this scoring method was found to predict accurately the severity of surgical stress in 89.4% cases. Discrepancies in the remaining cases were found to be related to specific clinical factors. On comparison of the hormonal and metabolic responses of neonates in the minor (N = 71), moderate (N = 12), and severe (N = 11) stress groups, increasing severity of surgical stress was found to be associated with greater and more prolonged changes in plasma catecholamines, blood glucose, and gluconeogenic substrates during and after operation. Clinical outcome following operation was also significantly different between the three stress groups.(ABSTRACT TRUNCATED AT 250 WORDS)