Recent advances in genetical engineering of the mouse have highlighted the importance of reproducible and less invasive models of cerebral ischemia in mice. In this paper, we developed minimally invasive and reproducible model of distal middle cerebral artery (MCA) occlusion in mice using krypton (Kr) laser-induced photothrombosis. C57BL/6 or BALB mice (n=8 each) were anesthetized with halothane. The skin was cut, the temporal muscle was retracted, and the right distal MCA was observed through the skull. A Kr laser beam of wavelength 568 nm was focused onto the MCA over the intact skull. Upon laser irradiation, intravenous administration of a rose bengal solution was begun. After 4 min of irradiation, the laser beam was refocused on the MCA just proximal to the first spot, and another 4-min irradiation was performed. Then, the right common carotid artery (CCA) was ligated. Three days later, the brain was removed, and infarct volume was determined. Infarction confined almost solely to the cortical area was produced in each mouse. Mean infarct volume in C57BL/6 mice was 25.2+/-13.7 mm3. The BALB mice group showed significantly larger and more reproducible infarction (44.1+/-5.2 mm3; the coefficient of variation was 12%) than did C57BL/6 mice (P<0.005). Our photothrombosis model of stroke in mice can be performed without craniectomy, and its reproducibility is satisfactory when using BALB mice.