There is good evidence that communities of practices (CoP) can impact systems around a common interest.^{(Reference Pyrko, Dörfler and Eden1-Reference Smart4)} In 2018 the National Nutrition Network–Early Childhood Education and Care (NNN-ECEC) was formed to promote healthy sustainable food environments in Early Childhood Education and Care settings, which aimed to positively impact child development and health outcomes through a systems lens and facilitating the translation of research in to practice and policy. Studies have highlighted the disconnect between the translation of research findings to practice and policy.^{(Reference Bornbaum, Kornas and Peirson5-Reference Ward, House and Hamer7)} To determine if the NNN-ECEC could add value to the sector knowledge transfer, several frameworks were considered in its’ establishment, maintenance and evaluation. These included a model to form CoPs,^{(Reference Wenger, McDermott and Snyder8-Reference Wenger10)} a socio-ecological model^{(Reference Kilanowski11)} and a task framework.^{(Reference Bornbaum, Kornas and Peirson5)} Data was collected at baseline and 12 months, which included network members self-identifying which knowledge brokerage tasks (quantitative) they had undertaken, supplemented by semi-structured interviews using Bornbaum and Kornas^{(Reference Bornbaum, Kornas and Peirson5)} task domains. Interpretation of the findings utilised a social learning model—Value Creation Cycles.^{(Reference Wenger, Trayner and De Laat9)} There was a clear shift in the tasks members undertook between the timepoints, which included members being more able to identify and obtain relevant information to support the networks endeavours. Members increasingly reached out to their own contacts to build the networks capacity. Barriers to sharing information in a competitive environment were apparent, so building trust in a new, more collaborative way of working was taking time. Reducing the focus on outputs and internal competitiveness increased members confidence to share information. The members grew to recognise such changes would require a much greater input that that from individuals or groups of individuals from an organisation, thus reducing these tensions. Clearly identifying and modelling respect for specific expertise, collegiality (and its benefits), built trust among members to ‘do things’ differently. It was clear members were comfortable with conventional networking to support tacit knowledge sharing, however it was evident this full value was not being realised and needed additional support from research and development experts to support new and better pathways of networking and knowledge exchange to truly impact systems for this sector in a positive way. The models cited in this study were used due to their historical application in CoP, and focused on individualistic measures, therefore having limited capacity to capture the value which was evident. Recommendations from this research suggest networks place emphasis on measuring social learning and knowledge brokering, to broaden the identification of the value of CoPs, through case studies and stories of impact. In addition to reaching out to ‘knowledge translation’ and ‘research and development’ experts to facilitate knowledge transfer from research, to practice to policy.