Good communication is of utmost importance in all forms of cancer care and especially so in the palliative context, where patients as well as relatives tend to be hypersensitive (Sandgren et al. 2010). To render the ungraspable graspable, met- aphors are frequently used drawing on their capacity to capture the intangible in terms of more familiar experiences. For instance, to die from cancer can be described as coming to the end of a life journey or losing a battle (Semino et al. 2015). The overarching goal of the project Metaphors in palliative cancer care (MEPAC), a Sweden-based three-year interdisciplinary research project involving linguists and health care researchers, is to strengthen the scienti c foundation for health care professionals’ understanding and use of metaphors in Swedish palliative cancer care. The project is inspired by the UK-based study Metaphor in end-of-life care (MELC). Our poster aims to give an overview of the entire project and present a snapshot of some preliminary ndings from a pilot study on blogs written by patients su ering from incurable cancer. Compared to other inter- net-based platforms such as chat rooms or discussion groups, ill-ness blogs are unique forums for self-expression. Personal blogs written by cancer patients have been observed to have the potential to contribute to nursing science’s body of knowledge and hence capability to alleviate the psychosocial burdens associated with cancer diagnosis (Heilferty 2009), which is why blogs were found particular- ly suitable for the current study. Furthermore, the Swedish blog arena stands out in international comparisons, because it is not delimited to young users but hosts a more varied range of writers (Andersson 2012). In addition to the blog data, the project also investigates interviews with patients, relatives and health care profes- sionals carried out within the frames of the Centre for Collaborative Palliative Care at Linnaeus University, Sweden. Our qualitative analysis of the blog data serves as a foundation for subsequent quantitative analyses using corpus tools in collaboration with the SWE-CLARIN initiative.