Aim To investigate the value of cranial computed tomography perfusion (CTP) imaging in evaluating conservative treatment response in patients with severe carotid artery stenosis (CAS). Methods A total of 90 patients with severe CAS were selected as the research subjects, all received conservative treatment, and they were divided into low response group and high response group according to whether cerebral ischemic disease occurred within 1 year. Regional cerebral blood volume (rCBV), regional cerebral blood flow (rCBF), regional mean transit time (rMTT), and regional time to peak (rTTP) were compared between the two groups. Cox regression model was used to analyze the independent risk factors of conservative treatment responsiveness, and nomogram was used to analyze the value of rCBV, rCBF, rTTP and rMTT in evaluating conservative treatment responsiveness, and the decision curve and clinical impact curve were verified. Results Of the 90 patients with severe CAS, 2 cases dropped out after 1 year of follow-up, and 88 cases were effectively followed up. There were 65 cases without cerebral ischemia-related complications (in the high response group) and 23 cases with cerebral ischemic diseases (in the low response group), including 15 cases of transient ischemic attacks (17.05%) and 8 cases of cerebral infarction (9.09%). After 3 months of treatment, the rCBF in the two groups was greater than that before treatment, and the rMTT and rTTP were lower than those before treatment (P＜0.05). Before treatment and 3 months after treatment, the rCBF of the low response group was lower than that of the high response group, and the rMTT and rTTP of the high response group were higher than those of the high response group (P＜0.05). The Cox regression model was used to screen out systolic blood pressure, uric acid, rCBF, rMTT and rTTP, and construct a nomogram prediction model for conservative treatment response in patients with severe CAS. The consistency index was 0.896. Calibration curve analysis showed that the prediction model predicted conservative treatment response was in good agreement with the actual conservative treatment response. Within a threshold range of 0.16 to 0.95, the net benefit rate of the combined nomogram model in assessing responsiveness to conservative treatment in patients with severe CAS was superior to testing alone. The number of people classified as high risk by the joint detection scheme and the number of true positive cases were basically the same when the threshold probability was 0.41. Conclusion CTP imaging parameters are closely related to treatment responsiveness in patients with severe CAS, and they can provide a reference for early clinical evaluation of treatment responsiveness and help ensure patient benefit.