This thesis argues that the management of environmental impacts has many analogies with the management of risk and practice can be improved by using ideas, concepts and principles found in the management of risk in other spheres of human activity. An overview of the challenges faced by environmental impact practitioners in New Zealand and reinforcement of its importance to the sustainable management of natural and physical resources under New Zealand's Resource Management Act is provided. Key risk management ideas, concepts and principles drawn from a variety of sources are listed and parallels drawn between these and existing environmental impact assessment practice in New Zealand. From this list a number are selected and opportunities for improving environmental impact assessment practice are explored. A number of opportunities are identified, starting with the need for a common language and methodology amongst practitioners. Categorisation of impacts to assist transparency of analysis and expression using frequency-consequence curves to aid and promote consistency of decision-making are further areas of opportunity. Risk management has several well-developed techniques for dealing with uncertainty and selection of assessment endpoints. The connection between communication of risk and public perception is an area with significant potential for communication about environmental impacts. Challenges with effective public participation in environmental decision-making are backgrounded and risk management practised in two high profile areas examined for opportunity for improved practice. Neither appears to offer opportunity for improvement in key decision-making areas. A relatively new indicator approach towards risk assessment called "healthy systems method" appears to have significant potential for cost-effective analysis of systems of various types and at various levels. This thesis identifies a number of other areas of risk management requiring further research to determine potential for achieving better practice in environmental impact assessment.