Despite the interest they generate in the public and media, changes of political party leadership in western democracies have received surprisingly little academic analysis. The existing knowledge and understanding of how and why party leadership changes is severely limited. New Zealand is no exception. In an attempt to rectify this situation, this study seeks to offer a theory which can sufficiently explain (and predict) leadership vulnerability of the New Zealand Labour Party. The emphasis is placed on caucus members who have exclusive power to select and de-select their leaders. Through two case studies - the 1993 leadership change and the unsuccessful 1996 leadership change attempt - against which the theory is tested, it is argued that quality of support that a leader receives on his/her selection has significant importance to her/his future.