Middle cerebral artery occlusion (MCAO) in rats is the most commonly used stroke model. Besides the infarct size, assessment of sensorimotor performance has become increasingly important in neuroprotective drug research. However, contradictions exist about procedures for testing functional outcome following MCAO. The aim of the present study was to evaluate a relatively simple set of neurological tests based on the most commonly used scoring systems, and to describe the functional recovery and correlation with the infarct size in rats sacrificed 2 or 14 days after permanent or transient MCAO. The smaller infarct size of rats with transient occlusion was reflected in the neurological scores only during the first 6h. By day 14, no recovery occurred in postural signs, lateral resistance and spontaneous activity, other signs showed different degrees of recovery. Correlation with the infarct size was found only on certain days in gait disturbance, placing reactions, daily body weight and spontaneous activity. According to our observations, the most commonly used sensorimotor tests provide a useful initial screening of functional deficits, but these tests most probably measure deficits caused by infarction of the core area. It is suggested that these tests should be completed by more refined tests when testing a neuroprotective drug which reduces the infarct size in penumbral areas.