The Mongolian gerbil has been widely used as a global brain ischemia model because of its incomplete cerebral circle of Willis. However, the inter-individual anatomic variability of this vascular structure interferes with the reliability of the model. The aim of this work was to introduce modifications to the protocol of global brain ischemia experiments in Mongolian gerbils in an attempt to increase the reliability and usefulness of this model. Our study focused on the assessment of the level of anastomosis of the cerebral circle of Willis in order to evaluate its contribution to clinicopathological outcomes in this model. Sham-operated, Ischemic, and Ischemic + Hypothermia animals were subjected to a 15-minute occlusion of the common carotid arteries. Transcardiac perfusion with bromophenol blue / gelatin solution was performed 72 hours after ischemia. Brains were processed for anatomopathological analysis. Tissue damage was observed in the hippocampus, caudate-putamen nucleus, neocortex, and thalamic nuclei of animals from the Ischemic group. The circles of Willis of the Sham-operated animals showed bilateral (38%), unilateral (48%) or no posterior communicating arteries (14%). A negative correlation between infarct volume and the level of anastomosis was revealed for the Ischemic, but not for the Ischemic + Hypothermia group. Additionally, Analysis of covariance (ANCOVA) was performed to assess the contribution of the level of anastomosis to the clinicopathological outcomes. It was confirmed that the infarct volume decreased in the Ischemic + Hypothermia group when compared to the Ischemic group. Since the level of anastomosis cannot be predicted, this variable should necessarily be considered when analyzing the results of global brain ischemia in Mongolian gerbils.