We aimed at investigating a new model of mild focal cerebral ischemia in rats with repeated, noninvasive magnetic resonance scanning combined with histology. Magnetic resonance imaging yielded information about infarct development enabling us to test the putative growth of the infarct over time. The effect of local temperature at the occlusion site in this model was furthermore tested. Thirty-three Wistar rats were subjected to 30 min of simultaneous common carotid artery and distal middle cerebral artery occlusion or sham treatment. Animals were magnetic resonance-scanned repeatedly between day one and day 14 after surgery, then sacrificed, and paraffin brain sections stained. All animals scanned 24 h after reperfusion showed an area of edema in the affected cortex, which later was identified as an infarct. Animals with a temperature of 33.9 +/- 1.5 degrees C at the MCA site (hypothermic) showed smaller infarcts (14.4 +/- 10 mm3) than animals with normothermic local temperature (36.7 +/- 0.2 degrees C, 57.7 +/- 26.4 mm3). Infarct size was maximal on day 3 after ischemia but decreased as edema subsided. Infarct volumes from histology and magnetic resonance imaging correlated well. The model reproducibly yielded cortical infarcts, which did not grow after edema had subsided. Local temperature had a considerable effect on final infarct size.