This study tested the speed and ease with which individual passengers could open and use the emergency exits currently provided on buses and coaches. The survey found that passengers believed that certain emergency exits, such as doors and hinged windows, would be easy to use in an emergency, that they knew how to use them and that instructions would help if they were uncertain. Twenty exit types were then tested with passengers, including emergency doors, ‘continental’ doors, roof hatches, hinged windows, breakglass windows and the emergency operation of service doors. It was found that passengers' expectations were generally not being met. Inappropriate handle design, location, feedback, anti-tamper cover guards and unclear instructions were found to result in significant delays in evacuation. The study recommends design improvements in these areas together with a consistent approach to the design and provision of emergency exit signage, conspicuity and instructions. It also recommends that high level doors are equipped with steps and that if window exits are to be retained these should be hinged rather than break-glass as the latter have severe disadvantages.