OBJECTIVE To explore the feasibility of optimized scan protocol in whole-brain vessel one-stop examination with 640-multislice computed tomography (640-MSCT) scanner. METHODS A total of 28 patients undergoing whole-brain vessel examination but showing no obvious cerebral disease with 640-MSCT scanner between September 2012 and May 2013 were collected and divided into two groups of A (n = 14) and B (n = 14) . The recommended scan protocol (protocol 1: collecting 19 volumes) was applied in A group while the optimized scan protocol (protocol 2: collecting 15 volumes) formulated by reducing scanning phases reasonably and changing collection intervals in B group. The dose length product (DLP) was recorded automatically and effective dose (E) measured. The CT perfusion (CTP) values and computed tomographic angiography (CTA) images were analyzed for both groups. The regions of interest (ROI) of CTP images with area (20 ± 2) mm² were located in bilateral frontal white matter, parietal white matter, centrum semiovate, basal ganglia, occipital lobe and cerebellum. The image quality of CTA was evaluated by two experienced radiologists using double-blind method. The results were analyzed by statistics. RESULTS Dose length product (DLP) in B group decreased 19.23% versus A group (3 419.40 vs 4 233.50 mGy·cm) .Every relative perfusion value of both sides from both groups were not statistically significant (P > 0.05) .Every relative perfusion parameter from individual territory in both groups showed no significant differences (P > 0.05) . The quality of CTA images between groups A and B were not statistically significant (P > 0.05) . CONCLUSIONS On the premise that the accuracy of perfusion parameters and the quality of CTA images, the optimized scan protocol in whole-brain vessel one-stop examination can obviously reduce radiation dose and it has important clinical significance.