The intraluminal filament model of middle cerebral artery occlusion (MCAO) in mice and rats has been plagued by inconsistency, owing in part to the multitude of variables requiring control. In this study we investigated the impact of several major variables on survival rate, lesion volume, neurological scores, cerebral blood flow (CBF) and body weight including filament width, time after reperfusion, occlusion time and the choice of surgical method. Using the Koizumi method, we found ischemic injury can be detected as early as 30 min after reperfusion, to a degree that is not statistically different from 24 h post-perfusion, using 2,3,5-Triphenyltetrazolium chloride (TTC) staining. We also found a distinct increase in total lesion volume with increasing occlusion time, with 30–45 min a critical time for the development of large, reproducible lesions. Furthermore, although we found no significant difference in total lesion volume generated by the Koizumi and Longa methods of MCAO, nor were survival rates appreciably different between the two at 4 h after reperfusion, the Longa method produces significantly greater reperfusion. Finally, we found no statistical evidence to support the exclusion of data from animals experiencing a CBF reduction of <70% in the MCA territory following MCAO, using laser-Doppler flowmetry. Instead we suggest the main usefulness of laser-Doppler flowmetry is for guiding filament placement and the identification of subarachnoid haemorrhages and premature reperfusion. In summary, this study provides detailed evaluation of the Koizumi method of intraluminal filament MCAO in mice and a direct comparison to the Longa method.