Networks are a significant conduit for the dissemination of innovations. However, van der Gaag and Snijders (2004) recognised that not all resources distributed throughout a social network would be as valuable in particular contexts as others. A resource generation model of social network analysis starts from the position that what matters for the adoption of an innovation - a change in working practice - is whether individuals are in contact with people possessing knowledge, capacity or other relevant resources that can assist the adopter with this change.This model forms the basis for a research project which aims at enhancing the resource generation networks within universities, vis-a-vis the dissemination of e-learning innovations. Many universities find this activity difficult. Universities can partly engineer resource generation networks by providing helpdesks, staff development units, and so on, but while these may create network hubs to an extent, there is still a need for the more personal, organic networks of resource generation (communities of practice, friendship links, and weaker ties such as mutual acquaintance, familiarity with prior work and so on). These are likely to extend outside the institution. Both the engineering and organic approaches are challenged by the structure of universities . Their fragmentation is now exacerbated by the development of institutions with widely separated satellite campuses, including in different countries.The project is currently in the pilot stage, and has mapped the social networks of two self-contained but fragmented institutions. The pilot has involved using a version of van der Gaag and Snijders??? research instrument to ask respondents who they would go to for help solving problems with e-learning. I seek to confirm the hypothesis that in these case organizations, there will be clear ???cut points??? in the network ??? individuals or small groups on whom the ability to use social networks to diffuse learning and innovation across the different parts of the organization will depend. Phase two will involve researching in a ???behavioural change intervention??? in one of the organizations, one targeted at key individuals or groups as defined above. Both organizations will then be re-examined after a year, checking the evolution of the resource generation networks over this period.REFERENCE:van der Gaag, M. and Snijders, T. (2004). Proposals for the Measurement of Individual Social Capital. In Flap, H. and Volker, B. (Eds.), Creation and Returns of Social Capital: A New Research Program. London: Routledge.