Several lines of evidence have demonstrated beneficial effects of the hematopoietic factor G-CSF in experimental stroke. A conclusive demonstration of this effect in G-CSF deficient mice is, however, lacking. We therefore investigated the effect of G-CSF deficiency on infarct volumes, functional recovery, mRNA and protein expression of the matrix metalloproteinase 9 (MMP-9) after stroke. Furthermore we tested the efficacy of G-CSF substitution in G-CSF deficient animals to prevent the potential consequences of G-CSF deficiency. In the present study experimental stroke was induced in female non-treated wildtype (wt), G-CSF deficient mice and G-CSF substituted G-CSF deficient mice followed by assessment of infarct volumes, neurological outcome and sensorimotor function. In addition, immunohistochemistry and real-time PCR of the peri-ischemic area were performed. G-CSF deficient mice showed increased infarct volumes, whereas G-CSF substituted mice had a remarkable reduction in lesion size compared to wt mice. These findings are accompanied by an improvement in neurological and sensorimotor function. G-CSF deficiency resulted in an upregulation of MMP-9 in the direct peri-ischemic tissue. Treatment with G-CSF suppressed the upregulation of MMP-9. Taken together, G-CSF deficiency clearly resulted in enlarged infarct volumes, and worsened neurological outcome. G-CSF substitution abolished these negative effects, led to significant reduced lesion volumes, and improved neurological outcome. G-CSF mediated suppression of MMP-9 further demonstrates that endogenous G-CSF plays a significant role in brain protective mechanisms. We have shown for the first time that endogenous G-CSF is required for brain recovery mechanisms after stroke.