The ever-increasing production of digital textual data in a wide variety of forms presents both opportunities and challenges to researchers. The opportunities derive from the rich availability of secondary data with which to work. The challenges are the familiar ones of lack of time and resources to undertake the analysis. The qualitative researcher is often faced with a trade-off between analysing a large amount of textual data at a cursory level and selecting (sampling) a small subset of that data for deeper analysis. This paper describes a new computer-aided technique for automatically analysing large amounts of textual data. The approach is proposed as an addition and a complement to the existing computer-aided qualitative data analysis software methods rather than as a substitute. The technique needs little investment either financially or in terms of learning to use the software. The particular software used has been developed principally for knowledge-management purposes; however it has clear potential for researchers. Examples of the use of the technique are given as well as suggestions for further applications. The paper also weighs the advantages and disadvantages of the technique compared with more familiar approaches. (Contains 16 notes and 8 figures.)