Objective To compare the effects of stent combined with drug therapy and aggressive drug therapy alone on cognitive function for patients with symptomatic severe intracranial vertebrobasilar artery stenosis and cognitive dysfunction. Methods A total of 53 patients from October 2016 to March 2019 with symptomatic severe intracranial vertebrobasilar artery stenosis and cognitive dysfunction were collected. Patients were divided into 2 groups [stent combined with drugs group (n = 31) and aggressive drugs alone group (n = 22)] based on patients' informed consent. Perioperative status of stent combined with drugs group and incidence of adverse events in both groups were analyzed. Montreal Congnitive Assessment (MoCA) scores were compared in hospitalization period, 1 month and 6 months at follow-up in order to evaluate the improvement of cognition. Then subgroup analysis of different collateral circulation status before stent combined with drugs group was conducted. Results There were 2 cases of perioperative complication in stent combined with drugs group, and there was no statistical difference in cumulative incidence of adverse events and ischemic stroke between 2 groups. According to the comparion that MoCA scores in hospitalization, 1 month and 6 months at follow-up in two different therapy groups, the score of delayed recall had statistical significance (P = 0.005), which indicted stent group had advantages in delayed recall over drugs group. The scores of total MoCA (P = 0.000), line connection test (P = 0.028), drawing clock (P = 0.006), attention (P = 0.001) and delayed recall (P = 0.000) had statistical significance in different follow-up time points and were significantly improved in stent group. In stent group, the scores of total MoCA (P = 0.000, 0.000, 0.009) and attention (P = 0.000, 0.000, 0.030) improved in succession in the time of before treatment, 1 month and 6 months at follow-up. However, the scores of line connection test (P = 0.001, 0.000), drawing clock (P = 0.006, 0.000), and delayed recall (P = 0.000, 0.000) improved mainly within first month at discharge. As to the scores of MoCA in different preoperative collateral circulation status groups, sentence repetition only had statistical significance (P = 0.021). The scores of total MoCA (P = 0.000), line connection test (P = 0.000), drawing clock (P = 0.002), attention (P = 0.000) and delayed recall (P = 0.000) had statistical significance in different follow-up time points. In further comparion of changes of MoCA scores at different time points in 2 preoperative collateral circulation status groups, scores of total MoCA (P = 0.006, 0.006) and attention (P = 0.005, 0.002) were higher at follow-up of 6 month than these in hospitalization and 1 month follow-up in good collateral group, which suggested the scores above improved mainly between 1 month and 6 month at discharge. The score of delayed recall at discharge after 6 month was higher than it in hospitalization. The scores of total MoCA (P = 0.000, 0.000), line connection test (P = 0.003, 0.001), drawing clock (P = 0.004, 0.001), attention (P = 0.000, 0.000) and delayed recall (P = 0.000, 0.000) were higher at follow-up of 1 month and 6 month than these in hospitalization in poor collateral group, which suggested the scores above improved mainly within first month at discharge. Conclusions As to the patients with symptomatic severe vertebrobasilar artery stenosis and cognitive dysfunction, the improvement in multiple cognitive domains of stent combined with drugs therapy and the advantage over aggressive drugs alone group were observed. Meanwhile, the time of cognitive benefit was earlier in poor preoperative collateral circulation status groups than in good collateral group. [ABSTRACT FROM AUTHOR]