Results from in-vitro experiments suggest that inhalational anaesthetics may have a detrimental effect on the course and incidence of Alzheimer's disease. However, case-control studies in humans show no negative impact of anaesthetics on the course of Alzheimer's disease. To test the hypothesis that 2 h of general anaesthesia with 1 MAC isoflurane changes learning abilities of young and old transgenic Alzheimer's mice (APP23 mice). Randomised controlled double-blinded study in mice. Animal laboratory and operating theatre in the Klinik für Anästhesiologie, Technische Universität München, Germany Ninety-six male mice divided in four groups: young (4 months) APP23 mice and corresponding wild-type mice; old (14 to 16 months) APP23 and corresponding wild-type mice. Mice were either anaesthetised for 2 h with 1 MAC isoflurane or sham-anaesthetised ('isoflurane' or 'control'). Learning and locomotor activity during the following 8 days using the modified Hole Board Test for mice. Results are median (interquartile range) and median difference (95% confidence interval). Young mice, [1.0 (1.3)] as assessed by the number of omission errors, learned better than old [1.8 (1.8); age: P = 0.004, median difference 0.5 (0.2 to 1.0)]. Anaesthetised animals [0.8 (1.5)] learned better than controls [1.6 (1.7); anaesthesia: P = 0.010, median difference 0.5 (0.1 to 0.9)]. This was accompanied by higher locomotor activity in young compared to old mice as assessed by number of line crossings per minute [10 (5) min(-1) vs. 7 (3) min(-1), P < 0.001, median difference 3 (2 to 4) min(-1)]. Anaesthesia and genotype Alzheimer's disease had no impact on locomotor activity. Isoflurane may have protective, rather than detrimental, effects on cognition in Alzheimer's disease.