Stroke is despite of progressive improvements in treatment and reperfusion strategies one of the most devastating human pathology. However, as quality of acute health care improves and more people survive ischemic attack, healthcare specialists have to solve new challenges to preserve reasonable quality of life to these patients. Thus, novel approaches which prevents comorbidities of stroke and improve quality of life of stroke survivors in general has to be developed and experimentally tested. The aim of the present paper was to establish reliable rat model of middle cerebral occlusion and set of methods allowing selection of animals suitable for long-term experiments. We have compared mortality rates, cerebral blood flow and extension of ischemic lesion induced by intraluminal filament in three widely used outbred rat strains. We have additionally used an animal 18F-DG PET scans to verify its reliability in noninvasive detection of ischemic infarct in acute period (24 h after MCAO) for selecting animals eligible for long survival experiments. Our data clearly indicates that high variability between rat strains might negatively influence stroke induction by intraluminal thread occlusion of middle cerebral artery. Most reliable outbred rat strain in our hands was Sprague-Dawley where maximal reduction of cerebral blood flow and extensive ischemic lesion was observed. Contrary, Wistar rats exhibited higher mortality and Long-Evans rats significantly smaller or no ischemic region in comparison to Sprague-Dawley. Additionally, we have confirmed a positron emission tomography with 18F-fluorodeoxyglucose as suitable method to assess extension of ischemic region in acute period after the experimental arterial occlusion in rats.