Undue intraoperative brain retraction can cause significant neurosurgical morbidity. By combining brain retractor blade pressure measurement with monitoring of brain electrical activity, one can determine the limits of safe brain retraction and then test systematically various therapeutic interventions. Cortical evoked potential (EP) mapping and laser-Doppler cerebral blood flow (CBF) measurement were undertaken during brain retraction in the miniature swine (Sus scrofaj. Forelimb somatosensory EP recording during subtemporal retraction simulated the pterional and subtemporal approaches, respectively. Retraction pressure of 30 mmHg usually resulted in a 50% decrement in EP amplitude after 10 to 20 minutes in normotensive, normocapnic adult animals. Recovery of EP occurred within 5 to 10 minutes of retraction release. The effects of animal age, induced hypotension (nitroprusside, MAP ≈ 40), and induced hypocapnia (hyperventilation, PaCO2 ≈ 28) on EP preservation during retraction were then investigated', with data reported here from 23 animals (8 to 35 kg). By Spearman rank correlation coefficientsearly loss of EP was associated with the following: lower MAP (p ≈ 0.0001), lower CBF (p ≈ 0.0005), lower PaCO2 (p < 0.001), and older age (p ≈ 0.01). These results indicate (1) retractors should be relaxed every 10-15 minutes whenever possible (for at least 5 minutes), and (2) hypotension, in particular, but also hypocapnia (hyperventilation) should not be used indiscriminately. Details of this new model of retraction ischaemia are presented.