Speeding is one of the principal causes of traffic accidents that harm not only the safety of pedestrians but also the driver himself and others. However, empirical studies on speeding behavior are mostly survey research and identification of cause-and-effect relationship is unclear. Also the reliability of the study results cannot be verified using retrospective methods. To complement this point, it is important to conduct an experimental study, but realworld field research using real vehicles is almost impossible to apply equally to all subjects by controlling surrounding vehicles and signal systems. Therefore, in this study, we compared the overspeed characteristics of high risk driver types using driving simulator. The high risk driver types were divided in to four types: elderly driver, commercial driver, driver with more than 10 years of driving experience, and driver within 2 years of license acquisition. The overspeed characteristics were identified by using an index called Accumulated Speeding (AS). Based on overspeed characteristics differences between driver types, it is expected that additional education by driver type will be possible in the safety education using Driving Simulator.