The purpose of the current study was to empirically examine the effects of accident scenes on eye movement as well as driving behavior. Fifty-four participants drove in a driving simulator wearing a head-mounted eye-tracker in three experimental drives, one of which had an accident scene. The participants were put into one of three different conditions (no barrier, partial barrier, or full barrier). The results showed significant main effects of distraction (accident vs. no accident) on dwell frequency and duration, average speed, and root mean square error of the steering wheel angle during the drive with the accident scenes. In addition, the results also showed significant interaction effects between distraction and type of barrier (no, partial, or full) on dwell frequency and duration. The full barrier condition had the biggest effect on decreasing dwell duration and frequency. The findings support the Salience Effort Expectancy Value (SEEV) model of attention and previous research stating objects high in salience attract attention (Wickens & Horrey, 2008; Itti & Koch, 2000). These findings also support previous research by Mayer, Caird, Milloy, Percival, & Ohlhauser (2010) stating that drivers drive in the safest manner (lowest passing speed) when an emergency vehicles are present with the emergency lights on. Temporary barriers could be used to help decrease the effects of rubbernecking on highways when an accident scene is present (Masinick & Teng, 2004; Potts, Harwood, Hutton, & Kinzel, 2010)