A large body of literature addresses the role of science in environmental impact assessment (EIA) but less attention has been given to the views of practitioners themselves. In this research a survey of 31 EIA practitioners in Western Australia was undertaken to determine their perceptions of the quality and importance of science in EIA. The survey results are compared with previous theoretical, empirical, and survey studies of the role of science in EIA. Interview questions addressed the role of science in impact prediction, monitoring activities, mitigation and management, and EIA decision-making. It was clear from the interviews that many practitioners are satisfied with the quality of science currently used in EIA, but do not believe that it is given sufficient importance in the process. The quality and importance of science in the predecision stages of EIA was rated higher than in the postdecision stages. While science was perceived to provide the basis for baseline data collection, impact prediction, and mitigation design, it was seen to be less important during decision-making and ongoing project management. Science was seen to be just one input to decision-makers along with other factors such as sociopolitical and economic considerations. While time and budget constraints were seen to limit the scientific integrity of EIA activities, pressure from the public and regulatory authorities increased it. Improving the scientific component of EIA will require consideration of all these factors, not just the technical issues.