Previous rat fMRI studies have been analyzed by statistical parametric mapping (SPM) but the SPM is designed for human fMRI studies. In this study, we examine whether the default settings including delay time, statistic methods and hemodynamic response function (HRF) shape in SPM can directly transform to rat fMRI studies. Results show statistic methods do not affect the mapping of brain activation. However, the delay time and HRF shapes have significant impacts on extracting brain activation, especially in primary somatosensory cortex and striatum. We therefore suggest the HRF should be optimized to approach better sensitivity in other rat fMRI studies.