Midazolam is a widely used sedative agent during colonoscopy, with cognitive toxicity. However, the potential cognitive hazard of midazolam-based light sedation has not been sufficiently examined. We aimed to examine the cognitive safety and vulnerability profile under midazolam light sedation, with a particular focus on individual variations. Methods: We conducted a prospective case-controlled study in an academic hospital. In total, 30 patients undergoing sedative colonoscopy as part of a health check-up were recruited. Neuropsychological testing on the full cognitive spectrum was evaluated at 15 minutes and 120 minutes after low-dose midazolam administration. The modified reliable change index (RCI) was used for intrapersonal comparisons and controlling for practice effects. Results: Midazolam affected psychomotor speed (48%), memory (40%), learning (32%), working memory (17%), and sustained attention (11%), while sparing orientation and the fluency aspect of executive function at the acute stage. Residual memory (10%) and learning (10%) impairments at 2 hours after administration were evidenced in some patients. The three object recall and digit symbol coding tests can serve as useful screening tools. Conclusion: Midazolam-based light sedation induced selective cognitive impairments and prolonged cognitive impairments occurred in patients with advanced age. A longer observation time and further screening were recommended for patients due to their at risk state.