Since the mid-1970s, some electoral analysts have observed the declining capacity of class-based party alignment to explain electoral outcomes in western democracies, including Australia. While this observation has never been entirely unchallenged, it is now broadly accepted within the discipline that while party alignment is still a significant factor, dealignment is gradually occurring. One consequence of the decline of party alignment is that the discipline of electoral analysis must seek new theories and models to explain the activity of those voters who do not vote in accordance with party alignment. A number of explanations have been advanced, including “issue voting”, based upon voters making rational judgments on contemporary policy issues; and “image voting”, based primarily upon the public relations image of the party and leader. One theory which has been advanced to explain a proportion of the vote in a limited number of Queensland electorates is connectional politics. Connectionalism occurs where a candidate, regardless of their party label or independence, obtains electoral support on a personal basis as a result of their position within, and “connections” within, the community. This position of leadership must be developed over a long time, and it is suggested that the strength of the relationship between the community leader and the community can, under some circumstances, be translated into electoral support.The concept was first described by Reynolds in Connectional Politics – The Queensland Case.1 This paper described a series of occurrences of connectional politics, but did not endeavour to explore the mechanics of connectional politics. It introduced the concept, but left the way open for further work to explore how, and under what circumstances, connectionalism can be effective. This thesis undertakes some of that further work, in an effort to describe the operations of connectional politics. This research extracts from Reynolds’ paper a series of “connectional indicators” related to the candidate, the community, and the electoral circumstances, then tests those indicators in five substantial case studies, observing potentially connectional candidates in various campaigns in the 1992, 1995, 1998 and 2001 state general elections. The result of this research is first, confirmation that connectional politics in Queensland does occur, supporting Reynolds’ initial contention. Second, however, this research identifies some important factors which contribute to a successful connectional campaign. The results of this research indicate that connectional politics occurs most readily in communities which consider themselves to be in need of instrumental services which must be provided by the central government. This sense of need is often coupled with a sense of neglect; a perception that the central government is neglecting the important needs of the community. In order for these needs to be realised, the community requires a “champion” whose primary purpose is to enter state politics in order to ensure that the state government provides support for the community’s self-identified needs. Such a “champion” is usually a person who is already a leader within the local community. They may, for instance, have served in local government, or taken leadership positions in local associations. This research suggests that if such a leader emerges, and if the core of his/her campaign is their desire to overcome the “neglect” shown by the central government, they can obtain significant electoral support regardless of their party label (or lack thereof). This conclusion is important, but carries a number of important caveats. The first is that this research is based upon five case studies. The studies were undertaken in some depth, and provide a stable foundation for the research conclusions, but they are still only five in number. This research will become stronger if studies such as these are undertaken by future researchers, during future elections. Second, the operation of connectionalism is constrained by circumstances. The community must have the necessary characteristics; the candidate must exhibit the necessary characteristics of leadership and community commitment; and the wider electoral circumstances must be amenable to the operation of connectional politics.