In the past few years, virtual worlds - such as Second Life (SL), World of Warcraft (WoW) and RuneScape - have demonstrated the potential to be a promising online business model. Millions of users around the world now participate in virtual worlds and trade virtual items with each other. What is more important is that typically virtual items can be traded for real world currencies within virtual worlds or via other third-party exchange platforms. The advent of virtual worlds, accompanied by virtual item transactions for real money has created a significant impact on the lives of people globally. However, little empirical research has been conducted into virtual world residents' purchase behaviour in virtual worlds. Based on the gaps found in the literature, this research was designed to help us to gain a better understanding of factors influencing virtual item purchase behaviour in virtual worlds. This research was carried out in two distinct stages, covering both exploratory and explanatory research. The exploratory stage began with reviewing the current stage of knowledge with respect to understanding players' purchase behaviour in virtual worlds. Seven prominent theoretical models were carefully selected as our theoretical frameworks to build a preliminary research model for explaining virtual item purchase behaviour in virtual worlds. The seven theoretical models are the theory of reasoned action (TRA) (Fishbein and Ajzen, 1975), the theory of planned behaviour (TPB) (Ajzen, 1991), the technology acceptance model (TAM) (Davis, 1989), trust theory (McKnight et al., 2002) transaction cost theory (Williamson, 1981, 1985, 1991; Liang and Huang, 1998), innovation diffusion theory (lOT) (Rogers, 1995) and the unified theory of acceptance and use of technology (UTAUT) (Venkatesh et al., 2(03). These established theoretical models have proven to have good predicting or explaining power via prior research in the field of IS adoption and e-commerce online shopping behaviour (e.g. Cheong and Park, 2005; Gefen, 2000; Hsu and Lu, 2004; Lin and Lu, 2000; Teo et al., 1999; Moon and Kim, 2001). Moreover, four focus groups were then conducted to gain deep insights into virtual item purchase behaviour in virtual worlds, thereby further refining and improving the preliminary conceptual model and its constructs. The revised research model provides a more comprehensive understanding of the latent psychological processes that typically induce players' purchasing behaviour in virtual worlds. For practical considerations, a more parsimonious research model was finally developed for empirical testing based on an overall modification and refmement of the revised model. In the explanatory research stage, a large-scale online survey was conducted within World of Warcraft and Second Life respectively. SEM-PLS (Structural Equation Models by Partial Least Squares) (Tenenhaus et aI., 2005) was used to confirm the finally developed model in this study. SEM-PLS can offer a unified mechanism to validate the relationships between constructs and their indicators in the measurement model and test the relationships among constructs in the structural model together. Among the statistically significant paths found in the final research model, three prominent IS adoption constructs including effort expectancy, performance expectancy and enjoyment still exerted considerable influence on virtual item purchase behavioural intention. Another typical construct habit not only had the direct effect on actual purchase behaviour in both SL and WoW contexts, but also significantly moderated the relationship between purchase behavioural intention and actual purchase behaviour in the context of WoW. Each of three newly developed constructs including advancement, perceived value and customisation had a strong impact on players' virtual item purchase behavioural intention in both WoW and SL contexts. As expected, perceived social status had a close relationship with players' purchase behavioural intention in WoW. In summary, the research findings provided useful information for virtual world developers and marketers in prioritising and allocating their resources to improve the impact of these constructs, all of which will ultimately stimulate players' purchase of virtual items. For example, given the important influence of enjoyment and customisation, the developers of a virtual world should carefully consider how to design novel, interesting contents and virtual items for simulating players' purchase behaviour. The underlying transaction mechanism must be designed and developed towards ease of use and high usefulness. In addition, game-oriented virtual world developers should refer to socially-oriented virtual world economic systems (e.g. Second Life) and consider establishing a currency exchange mechanism for virtual currency and real money. This PhD dissertation rounds off with a discussion about future research opportunities, the limitations and implications for research and practice in this very new area of investigation.