Sevoflurane is widely used in anaesthetic protocols for patients undergoing surgical procedures. However, there are no reports on the influence of sepsis on minimum alveolar concentration of sevoflurane (MAC(SEV)) in animals or in humans. The aim of this study was to test the hypothesis that sepsis could alter the MAC(SEV) in a normotensive septic pig model. Twenty young, healthy pigs were used. After they had received 10 mg kg(-1) of ketamine i.m. for premedication, anaesthesia was established with propofol 3 mg kg(-1) and the trachea was intubated. Sevoflurane was used as the sole anaesthetic agent. Baseline haemodynamic recording included electrocardiography, carotid artery blood pressure and a pulmonary thermodilution catheter. Baseline MAC(SEV) in each pig was evaluated by pinching with a haemostat applied for 1 min to a rear dewclaw. MAC(SEV) was determined using incremental changes in sevoflurane concentration until purposeful movement appeared. Pigs were assigned randomly to two groups: the saline group (n = 10) received a 1-h i.v. infusion of sterile saline solution while the sepsis group (n = 10) received a 1-h i.v. infusion of live Pseudomonas aeruginosa. Epinephrine and hydroxyethylstarch were used to maintain normotensive and normovolemic haemodynamic status. In both groups, MAC(SEV) was evaluated 5 h after infusion. Significant increases in mean artery pulmonary pressure, filling, epinephrine and vascular pulmonary resistances occurred in the sepsis group. MAC(SEV) for the saline group was 2.4% [95% confidence interval (CI) 2.1-2.55%] and the MAC(SEV) for the sepsis group was 1.35% (95% CI 1.2-1.45%, P<0.05). These data indicate that MAC(SEV) is significantly decreased in this normotensive septic pig model.