In laboratory animals as well as in human beings a depth of anaesthesia, where the subject has no pain or recall of events from the surgery, should be provided. Haemodynamic parameters such as heart rate and blood pressure are not a guarantee for an optimal depth of anaesthesia, especially when using neuromuscular blocking agents (NMBA). A number of studies suggest that the Middle Latency Auditory Evoked Potentials (MLAEP) contain information about the state of consciousness in humans. The purpose of this study was to examine whether the AEP could serve as an indicator of depth of anaesthesia in rats. The AEP was elicited with a click stimulus and monitored in an 80 ms window synchronised to the stimulus. The AEP was extracted applying an Auto Regressive Model with Exogenous Input (ARX-model) from which a Depth of Anaesthesia Index (DAI) was calculated. DAI was normalised to 100 while awake and decreasing gradually to a level between 50 and 20 as the rat was anaesthetised. Nine rats were anaesthetised and included in the study. Four doses of Hypnorm vet. and Dormicum were given as a total, each with 5 minutes interval. Clinical signs of the level of anaesthesia were observed simultaneously with the AEP. The results showed that in four rats DAI decreased to a level below 30 while anaesthetised. In the remaining five rats the AEP was only decreased to a level below 45. The results indicated that a simple dosing regimen based on weight was unable to give the same depth of anaesthesia in individual rats. The decrease in the DAI correlated well with the loss of stimulus response. In conclusion, MLAEP could be used as an indicator of depth of anaesthesia in rats during Hypnorm vet. and Dormicum administration. However studies applying other anaesthetic drugs should be carried out, before a conclusion of the general utility of the method can be made.