GH treatment can increase the mortality and morbidity of critically ill patients. The mechanisms of these harmful effects of GH are unknown but have been, in part, ascribed to interactions between GH and the immune system. Because GH has pattern-dependent actions we have now compared the dose-related effects of continuous and intermittent GH treatment given with or without an endotoxin (lipopolysaccharide; LPS) challenge. Male Wistar rats (n=6 per group) were treated for 5 days with recombinant human GH (0, 10, 100 or 1000 microg/kg per day) using either continuous s.c. infusion by osmotic minipump or intermittent twice daily s.c. injections. On day 4, endotoxin (5 mg/kg, i.p.) was injected and the animals monitored for a further 16 h. LPS administration alone led to neutrophilia and lymphopoenia, with increased plasma concentrations of urea, cholesterol, triglyceride, insulin and leptin, and decreased levels of IGF-I. High dose GH infusion (1000 microg/kg per day) followed by LPS caused greater increases in plasma urea, cholesterol, triglyceride, sodium and magnesium, but lower plasma glucose and insulin levels, than treatment with LPS alone. In contrast, twice daily injections of GH did not enhance these effects of endotoxin. In conclusion, the effects of endotoxin on plasma electrolytes, lipids, urea, glucose and insulin are differentially affected by the pattern of GH administration in the rat.