Ischemia models are indispensable for the evaluation of measures to be clinically applied to brain trauma or stroke patients. Slice models provide good control over experimental parameters and allow for comparative examinations of human and animal brain tissue. Experimental tissue, however, may be altered by anaesthesia, preparatory technique, and, in the case of human tissue, by underlying diseases. These influences on tissue behaviour under ischemia were examined electrophysiologically. Native rat tissue slices were prepared either immediately after decapitation (n = 13), during short ether/barbiturate narcosis (n = 18), or after two hours of inhalation anaesthesia (n = 12) imitating clinical narcosis. Tissue from rats in which generalized amygdala-kindled seizures had been triggered by electric stimulation (n = 10) was prepared according to the decapitation protocol, while human tissue (n = 10) was obtained during epilepsy or tumour surgery. Electrophysiological data (latency and amplitude of anoxic depolarization, recovery of evoked potentials) were recorded during ischemia simulation. Neither details of preparation or anaesthesia nor a history of epileptic fits were associated with significant changes of electrophysiological reactions under ischemia. Human tissue showed a significantly higher ability to uphold transmembrane ion gradients under ischemia. The ability of brain tissue to withstand ischemia is obviously species dependent. For the transfer of experimental results into clinical use it is important that interspecies differences alone can bring about a significant change of tissue behaviour.