We have studied the interaction of pregnenolone sulphate and pregnenolone with 5 beta-pregnanolone- and hexobarbitone-induced anaesthesia in male rats using an EEG threshold method. Burst suppression of the EEG of 1 s or more ("silent second" (SS)), was used as a criterion of deep anaesthesia. The effects of the steroid solvents albumin and beta-cyclodextrin were assessed by dose-response curves. Despite a significant increase in hexobarbitone threshold dose in relation to increased doses of albumin, there was no correlation between albumin dose and hexobarbitone concentrations in serum, fat and brain tissues. There was no significant difference in threshold concentrations of hexobarbitone between controls given albumin and those pretreated with pregnenolone. In subsequent experiments, 20% beta-cyclodextrin was used as steroid solvent and its volume was maintained at less than 3.0 ml kg-1 during pretreatment. Neither pregnenolone sulphate nor pregnenolone significantly altered the potency of 5 beta-pregnanolone for induction of anaesthesia. Furthermore, there was no interaction of pregnenolone sulphate and pregnenolone on induction of anaesthesia when hexobarbitone was used for anaesthesia.