OBJECTIVE Despite improved supportive care, the mortality of sepsis and septic shock is still high. Multiple changes in the neuroendocrine systems, at least in part, are responsible for the high morbidity and mortality. A reduced circulating level of insulin-like growth factor and an elevated level of growth hormone are the reported characteristic findings early in the course of sepsis and septic shock in adults. The aim of this study was to evaluate the changes of growth hormone/insulin-like growth factor 1 axis in sepsis and septic shock and investigate the relationship between these hormones and survival. METHODS Fifty-one children with sepsis (S), 21 children with septic shock (SS) and 30 healthy, age- and sex-matched children (C) were enrolled in this study. Growth hormone, insulin-like growth factor 1 and cortisol levels of the sepsis and septic shock groups were obtained before administration of any inotropic agent. RESULTS Growth hormone levels were 32.3 +/- 1.5 microIU/mL (range 4-56), 15.9 +/- 0.6 microIU/mL (range 11-28) and 55.7 +/- 2.7 microIU/mL (range 20-70) in S, C and SS groups, respectively. The difference between the growth hormone levels of the S and C groups, S and SS groups, and C and SS groups were significant (P < 0.001). Non-survivors (54.7 +/- 1.6 microIU/mL) had significantly higher growth hormone levels than survivors (29.4 +/- 1.5 microIU/mL) (P < 0.001). Insulin-like growth factor 1 levels were 38.1 +/- 2.1 ng/mL (range 19-100), 122.9 +/- 9.6 ng/mL (range 48-250) and 22.2 +/- 1.9 ng/mL (range 10-46) in the S, C and SS groups, respectively, and the difference between the insulin-like growth factor 1 levels of the S and C, S and SS, and C and SS groups were significant (P < 0.001). Non-survivors (8.8 +/- 1.1 micro g/dL) had significantly lower cortisol levels than survivors (40.9 +/- 2.1 microg/dL) (P < 0.001). We detected a significant difference between the levels of cortisol in non-survivors (19.7 +/- 1.8 microg/dL) and survivors (33.9 +/- 0.9 microg/dL) (P < 0.01). CONCLUSIONS There were elevated levels of growth hormone with decreased levels of insulin-like growth factor 1 in children during sepsis and septic shock compared to healthy subjects. In addition, there were even higher levels of growth hormone and lower levels of insulin-like growth factor 1 in non-survivors than in survivors. We think that both growth hormone and insulin-like growth factor 1 may have potential prognostic value to serve as a marker in bacterial sepsis and septic shock in children. As there is insufficient data in the paediatric age group, more studies including large numbers of patients and additionally evaluating cytokines and insulin-like growth factor binding proteins are needed.