The effects of the fentanyl fluanisone combination (Hypnorm(R)) and pentobarbitone sodium (Pentobarbital(R)) anaesthesia on blood glucose, insulin and glucagon were tested in rats in the fed and fasted state. Blood glucose was measured before and at 10, 20 and 30 min after injection of the anaesthetic agents. At 30 min the rats were sacrificed, and blood was drawn for measurement of glucagon and insulin. Pre-anaesthetic values for insulin and glucagon were established in separate groups of fasted and fed rats. In fasting rats given Hypnorm(R), blood glucose and plasma insulin were unchanged while there was a non-significant increase in plasma glucagon. The fasted rats given Pentobarbital(R) had unchanged blood glucose and plasma insulin and a non-significant depression of glucagon. The fed rats given Hypnorm(R) had a significant increase in blood glucose at 10 min and nearly a doubling of glucose values at 20 and 30 min (P < 0.001). Glucagon increased far less than in the fasted group, whereas insulin was doubled from preanaesthetic values (P < 0.05). The fed rats given Pentobarbital(R), had unchanged blood glucose, a slight non-significant depression of glucagon and a significant increase in insulin (P < 0.01). Thus Hypnorm(R) induced hyperglycaemia in fed but not in fasted rats, probably because more glucose was available in the fed state. Fed animals are a modification of the standard fasted animal model, and may be preferable when exploring hyperglycaemic or other reactions to anaesthetic agents.