Information on the equine stress response to anaesthesia and surgery is sparse but offers a promising approach to elucidating the high anaesthetic risk in this species. Previous work has shown that halothane anaesthesia induces substantial metabolic and endocrine changes. This paper reports the effects of barbiturate anaesthesia. Anaesthesia was induced with thiopentone in six ponies and no further agents were given. They stood within 30 mins. On another occasion, these animals, and three further ponies, were anaesthetised with pentobarbitone and anaesthesia was maintained for 2 h. No surgery was performed on either occasion. Plasma concentrations of glucose, lactate, non esterified fatty acids, cortisol, insulin, catecholamines and adrenocorticotrophic hormone were measured at the same time intervals in both groups before, during and after anaesthesia. There were no significant changes in hormones or metabolites during either period of anaesthesia and normotension was maintained. This was in marked contrast to the substantial stress response and hypotension under halothane anaesthesia in the same ponies. These results suggest that barbiturates may induce less of a stress response than halothane in horses. Recovery after 2 h of pentobarbitone anaesthesia was poor, precluding its clinical use. The need for a non-cumulative intravenous agent or a non-hypotensive volatile agent for use in equine anaesthesia is discussed.