Abstract The accuracy at which changes in cortical functional topology can be assessed by functional MRI (fMRI) depends on the quality of the reference coordinate system used for comparison of data sets obtained in different imaging sessions. Current procedures comprise an overlay of activation clusters on registered high-resolution anatomical images. Yet, fMRI images are frequently distorted due to susceptibility artifacts, which are prominent in rodent studies due to the small dimensions involved and high magnetic field strengths used. Therefore, a procedure for co-registration of activation maps has been developed based on anatomical landmarks defined on fMR echo planar images (EPI) themselves. Validation studies in control rats revealed that the centers of activated areas in somatosensory cortex S1, evoked through sensory forepaw stimulation fell within an area of 1 × 1 mm 2 in agreement with known electrophysiological coordinates. The technique was applied to detect changes in activation patterns in rats following smaller unilateral spinal cord injuries (SCI) in their cervical segments (C3/C4) 12 weeks after lesion. Despite of an almost complete behavioral recovery, fMRI responses remained altered in SCI animals with both significantly reduced fMRI signal amplitude and reduced latency to reach the peak response. Moreover, in SCI animals the activated S1 area corresponding to the contralesional forepaw was significantly enlarged and the center-of-mass for the ipsilesional paw was shifted rostrally. The mapping technique described combined with the temporal analysis of the BOLD response enabled a noninvasive quantitative characterization of cortical functional reorganization following SCI in rats.