AbstractFocusing on a specific vulnerable group (witnesses to crimes who have experienced or fear intimidation from offenders and their associates), this paper examines attempts by policy‐makers to address their anxieties and fears, and the impact of experiencing intimidation on the witnesses themselves. The first section traces the ‘discovery’ by policy‐makers of vulnerable witnesses in Britain in the mid‐1990s and how this led to the introduction of a range of spatial tactics and strategies aimed at reducing the vulnerability of witnesses to intimidation. This policy response has, however, been founded on a particular reading of the geography of witness vulnerability, narrowly focused around the courtroom during the trial period. Using interviews with witnesses who have experienced intimidation, the second section examines how the spatial and temporal coordinates of witnesses' fears are more complex than current policy measures allow for. The third section of the paper then juxtaposes policy and personal discourses around intimidation, focusing on the different understandings of the social, spatial and temporal contexts of intimidation and the differences between situational and social approaches to addressing witnesses' concerns. Many of the issues raised in this section concerning the dangers of relocating risk as a result of adopting situational approaches to witness vulnerability, and the tendency to address symptoms rather than causes, intersect with wider concerns that geographers have in relation to vulnerability research; these are discussed in the conclusion. Copyright © 2005 John Wiley & Sons, Ltd.