Röling and others argue that research and extension should not be seen as separate processes involving distinct institutions which must somehow be linked. Rather, scientists, specialists, extension workers, consultants and producers should be seen as participants in a single agricultural knowledge and information system (AKIS). This perspective on research, extension and adoption as activities that occur within a network offers new insights into the way in which technology transfer occurs, and new ideas as to how to increase the rate of adoption of new ideas and techniques in farming. In this study, we use network management theory and applied network analysis to operationalise the concept of an AKIS and to test the relationships between system characteristics and the effectiveness of technology transfer. The paper is based on the analysis of two networks associated with a large multi-site regional research and extension project in the temperate high rainfall zone of south-eastern Australia. The results indicate that network structure and the performance of individual network members have a significant impact on the ability of an AKIS to facilitate technology transfer.