Objective To explore the value of transcranial Doppler (TCD) hemodynamic parameters in predicting decompensated intracranial hypertension after craniocerebral injury. Methods A total of 122 patients with craniocerebral injury were monitored by TCD once a day for up to 7 days. The detection would be stopped if the patient showed decompensation. According to the results of cranial hypertension, patients were included in the positive group (n=35) and negative group (n=87). The hemodynamic parameters of MCA, such as systolic maximum blood flow velocity (VS), diastolic maximum blood flow velocity (VD), mean blood flow velocity (VM), pulsation index (PI), impedance index (RI), and the changes of parameters △Vs、△Vd、△Vm、△PI and △RI were recorded. The hemodynamic parameters of the positive and negative groups were compared, and the trend of hemodynamic parameters of MCA was analyzed. Cox regression analysis was used to reveal the risk of decompensated intracranial hypertension. ROC curve was used to analyze the accuracy of predicting decompensated intracranial hypertension. Results The incidence of decompensated intracranial hypertension in 122 patients with craniocerebral injury was about 28%. In the negative group, Vs, Vm and Vd showed a downward trend in the monitoring period, while RI and PI showed an upward trend. The max (△PI) of positive group was higher than that of the negative group, and the difference was statistically significant (P = 0.014). Max (△PI) can independently indicate the risk of decompensated intracranial hypertension (HR=1.71, P=0.027). The AUC of predicting decompensated intracranial hypertension was 0.904. The best diagnostic value was 0.148, and the sensitivity and specificity was 80.00% and 90.80%, respectively. Conclusion Using TCD to monitor and calculate max (△PI) can more accurately predict the probability of decompensated intracranial hypertension in patients who are still in compensatory period after craniocerebral injury.