The study reported on here focuses on self-efficacy in relation to high-school teachers’ teaching of writing. 140 New Zealand teachers from four schools completed a teacher-of- writing self-efficacy scale (TWSES) based on a rhetorical model of the writing process and incorporating five hypothesized dimensions. An initial principal components analysis was undertaken on 25 individual self-efficacy items to investigate the dimensionality of the data and the extent to which it reflected the dimensions hypothesized. A two-component solution emerged, termed “pre-writing instructional strategies” (accounting for 52% of total variance) and “compositional strategy demonstration” (7% of variance). Further principal components analyses conducted on groups of items deemed to be thematically coherent, that loaded on each component, confirmed that the data set for each group, treated sepa- rately to any other items, was approximately uni-dimensional. Measurement scales were calibrated to each group of items, and served as the dependent variables for comparisons of teachers’ self-efficacy in different subjects. Statistically significant variations occurred in the resultant scale locations for teachers of English, the humanities, science and mathe- matics. The study findings have implications for the teaching of writing as conceptualized in the secondary school, and indicate a value in viewing disciplinary literacies in rhetorical terms.