This paper seeks to review and examine the major theoretical underpinnings of research into power relations in IS projects. In order to keep the review manageable and to furnish a reasonable explanation of the ideas and papers referenced, the review is restricted to the ideas of Foucault, Giddens and Clegg. Partly, this restriction was due to the wish of the authors to deal seriously with the ideas presented, and not to produce an exhaustive but terse and taxonomic style review. Partly, the selection of the above theorists is due to the fact that the major IS papers on power relations have drawn on their work. This paper concludes with the presentation of a new theory of power and social influence from social psychology suitable for use in IS research. The ideas of this theory have not yet been examined empirically. Thus, IS researchers are challenged to employ these ideas in their empirical research.