Minimally invasive functional near infrared spectroscopy (fNIRs) technology was utilized to assess the effects of mannitol and hypertonic saline (HS) in treating traumatic brain edema (TBE). Rats with TBE models were given mannitol or HS in different dosages for different groups. The reduced scattering coefficient ( ${\mu }_s^{\prime }$ ) of the local cortex of rats was simultaneously monitored and recorded in vivo and real time by the minimally invasive fNIRs system. Brain water content (BWC) was measured by the wet and dry weight method at 1, 6, 24, 72, and $120\mathrm{h}$ after injury and treatment. Effects of treating TBE with different dehydration agents were then assessed by recording ${\mu }_s^{\prime }$ and BWC before and after administration of dehydration. In this study, the dynamic changes of brain edema and the effects of dehydration therapy were continuously monitored. Results implied that ${\mu }_s^{\prime }$ of the local cortex in rats is a good indicator for assessing effects of treatment of TBE. By recording changes in the value of ${\mu }_s^{\prime }$ , the following conclusions were obtained: HS is more effective than mannitol in reducing cerebral edema. The effect of dehydration of HS is only related to osmotic gradient and has no correlation with concentration.