Cerebral ischemia is a common sudden cerebral surgical disease with a high lethality and disability rate. The rapid and accurate detection of cerebral ischemia is of great significance to the diagnosis and treatment of cerebral ischemia. Inthispaper, we performed spectroscopy on cerebrospinal fluid (CSF) and serum of rats with ischemia time of 0, 0.5, 1, 2, 4, 6 and 24 h respectively, based on attenuated total reflection terahertz time-domain spectroscopy (THz-TDS). The changes in absorption coefficient and refractive index of CSF and serum with different ischemic times were analyzed. The results showed that the absorption coefficient and refractive index of CSF and serum of rats with different ischemic times were somewhat different compared with the control group. Furthermore, according to the absorption coefficient of CSF and serum with different ischemic times, principal component analysis and machine learning algorithms were used to automatically classify and recognize the degree of cerebral ischemia in rats. Especially, the recognition accuracy of the support vector machine classification model based on the absorption coefficient of CSF is relatively high, reaching 89.3%. Combining terahertz spectroscopic detection of CSF and serum ofrats with machine learning algorithms provides a new and effective detection method for the early diagnosis of cerebral ischemia.