Blogging is a recent phenomenon with research currently focusing on how it facilitates both personal and organisational knowledge exchange (Aimeur, Brassard & Paquet 2005; Hsu & Lin 2008). Social capital is shown to be a crucial factor facilitating knowledge transfer (Nahapiet and Ghoshal 1998). Blogging is a new social communication technology enabling individuals to collaborate and share knowledge. This research investigates how three dimensions of social capital affect individual knowledge sharing in weblog communities. In particular, it explores how individuals exploit weblogs as a tool for conversational knowledge management in educational institutions. Following Wasko & Faraj's (2005) study, the conceptual model is developed by setting eight independent variables from social capital dimensions and a dependent variable is set from individual behaviour in online knowledge sharing. Eight hypotheses are developed to test the relationship between these variables. A quantitative approach was applied for data collection and analysis. For data collection, an online survey was published in several Australian university weblog communities. An additional paper-based survey was distributed to the respondents in order to gain adequate sample size. For data analysis, confirmatory factor analysis (CFA) was applied to eliminate measurement items that shared a significant residual value with other measurement items. Further, the models obtained from confirmatory factor analysis were used to test the hypotheses by multiple regression analysis. Results from multiple regression analysis on online knowledge sharing suggest that trust, personal reputation and enjoy helping are positively associated with individual online knowledge sharing. The stepwise estimation procedure was further adapted in the regression model. The results show that four independent variables became significant to the study. These four significant variables were individual expertise, trust, personal reputation and enjoy helping. Lastly, several limitations in this study such as the sample of university online setting and respondents' activities on weblogs are discussed. These limitations lead to the direction of future research provided in conclusion of this study.