Within the work-life research and the debate between integration and segmentation (Nippert-Eng, 1996), the role of work/non-work preferences is central as a shift from segmentation to integration is often described in our contemporary society. It is often argued that organisations need to adapt policies to these different levels of preferences thus avoiding the duality of the integration versus segmentation dilemma. Such alignment shall avoid misfit and lower work/non-work conflict. The role of preference shall thus be explored. First, the existence of preferences is more assumed (see Ashforth, Kreiner, & Fugate, 2000; Kossek, Lautsch, & Eaton, 2005) than empirically demonstrated. Rothbard et al. (2005) explore such continuum but still fail to look at the continuum in its diversity. They indeed only measure the desire for segmentation and conduct their analysis considering segmentors versus integrators. Additionally, the relations with the segmentation-integration continuum and the work/non-work preferences have not been clearly made whereas the relation between boundary management and the continuum have been made (see Bulger, Matthews, & Hoffman, 2007). Second, the degree of explicitly of preferences is also not clear. Whereas it seems to be essential to know an individual’s preference while considering a proactive boundary work, Kossek et al. (2005, p. 351) indicate that “everyone has a preferred, even if implicit, approach for meshing work and family roles”. The roles of explicitness need to be made explicit where its implications on the nature of the process may be essential to understand an individual’s work/non-work experiences. Third, the conflicting results on the relation between preferences and enactment of boundaries question the process through which individuals evaluate their work/non-work preferences. Indeed, whereas conflict and balance were solely not enough and conceptually not relevant to understand individuals’ work/non-work experiences, they could make sense when associated with preference and the notion of fit. A misfit leads to conflict and a fit to balance. But if one can enact boundaries that are not based on preferences, where does that lead individuals? I am thus convinced that it is essential to go back to the origins of preference beyond what has been advanced so far. As a matter of fact, more just than being assumed, preferences are rarely explained in regards to individuals’ work/non-work experiences. Ashforth et al. (2000) indicate that individuals would choose to segment in order to minimize blurring of roles and to integrate in order to facilitate boundary role crossing. Such a view has recently been supported by de Man (2008). However, this explanation is not as revealing about why such preferences are indeed experienced by individuals. Back to cognitive behaviour (Zerubavel, 1991) and Nippert-Eng (1996), one can understand that social mindsets and one’s cultural environment may affect what could be called ‘preference’. In that regards, preferences are socially constructed. As a social constructionist, I share this view. However, it still stops short of explaining in practical terms why individuals’ experience one preference rather than another. As a whole, there is a necessity to understand work/non-work preferences on the integration-segmentation continuum, their level of explicitness, and their origins. The aim of this paper is to explore individual's work/non-work preferences between integration and segmentation as well as their level of explicitness and their origins. This paper is based on a narrative study of middle-managers. The study took place over a year. Data were generated through a structured series of open interviews as well as qualitative diaries. The aim of the collection process data was to uncover individuals' life domains and individuals' work boundary and boundary management processes including their work/non-work preferences. The analysis of the narratives enables to underline the preferences' explicitness and their origins. First, the research empirically demonstrates that individuals have preferences. It shows that there are two level of preferences: an overall level and a preferences in regards to the different elements that one can integrate and/or segment such as time, space, emotions, thoughts, individuals or even behaviours. Based on these two levels of preferences, an explicitness scale is defined. In turn, categories for work/non-work preferences are discussed so that a new classification of 5 preferences is presented: pushed segmentation, moderate segmentation, balanced segmentation, moderated integration and pushed integration. Second the results reveal that preferences are social and have four main origins namely 1- the work/non-work related event in one’s upbringing, 2-one’s past work/non-work experiences, 3-one’s understanding of one’s roles in their life domain and 4- one’s sense of personality and sense of “oneself".