Outdoor education involves the interaction of people with the natural environment, often in challenging situations. Because of this there are often real risks involved that must be suitably managed. Despite efforts at managing these risks, incidents are still occurring during outdoor education experiences that sometimes culminate in serious injury and even death. While in the past these injuries and fatalities would have been considered unfortunate acts of misadventure, new attitudes in society seek to ascribe blame in the wake of an accident and those involved, or their families, seek penalties for those blamed. Recent legislation such as the Health and Safety in Employment Act make these penalties easier to apply, while the recent actions of both the police and officers of the Department of Labour show the willingness of public officials to investigate and prosecute outdoor education providers if accidents occur. The outdoor education sector has been poorly equipped to reply to the public in the wake of serious incidents. There are few if any statistics on incident rates in the outdoor education sector, and there is very little known about the underlying causes of those incidents. To ensure that outdoor education provision can continue into the future such information needs to be available and training programmes developed based on those findings. This research builds a profile of almost 2000 incidents that occurred in the years 1996 - 2000 at 12 of the 25 larger outdoor education centres in New Zealand that were invited to participate. That profile includes calculating accident rates for the group of organisations sampled and compares these to the rates of accidents occurring in outdoor centres in other parts of the world as well as those occurring in other aspects of life in New Zealand. Eighteen of the incidents were chosen that had potential for serious injury, and these were studied for the root causes of the incidents using a Delphi technique involving three panels of outdoor experts. From this investigation, and an in-depth review of literature from the fields of safety management and psychology, I developed a taxonomy of root causes of outdoor education incidents and suggest a new model of how these root causes can interact to result in an incident. Not all of the identified types of error in the taxonomy of root causes could easily be accommodated within the existing frameworks of outdoor decision-making. In order to provide a model that incorporated these error types, theories of cognitive psychology were combined into a new model of outdoor education decision making in hazardous situations. This shows how personality factors, attitudes and other social factors can act to bias decisions and lead to incidents occurring. As a result of this research, an ongoing collection of incident data in the outdoor education sector is advocated, as is the adoption of the taxonomy of root causes and model of an outdoor education incident into training programmes for outdoor instructors. Through these processes it is hoped that risk management practices will be improved, incidents reduced in frequency and severity, and therefore participation of young people in outdoor education programmes for personal development outcomes can continue to be promoted and justified. Suggestions for further research to build the knowledge of the processes leading to incidents in outdoor education activities are made at the end of this thesis.