What works for whom but shouldn’t and what doesn’t work for whom but should?

Reviews of research evidence suggest that specific types of therapy are likely to be useful in treating particular types of disorder. Thus, the present commitment to evidence-based practice may mean that clients are automatically allocated to the treatment of choice according to formal research evaluation, purely on the basis of their presenting problems. This case study of two female clients referred for psychological therapy for trichotillomania illustrates the need to consider clients' general functioning and wider social context when decisions are made regarding the most appropriate mode of intervention. Contrary to expectations in accordance with research evidence, the client who participated in psychodynamic therapy had a successful outcome in terms of dramatic symptom reduction and improvement in general functioning. This client's outcome was, however, consistent with predictions made in the light of information from her repertory grid and her scores on measures of personal style. The client who participated in cognitive-behaviour therapy dropped out of therapy after five sessions. Clues as to why she may have done so are apparent both from her repertory grid, and from the transcript of her fifth therapy session. This paper illustrates the utility of the repertory grid in identifying aspects of client functioning which may be useful in matching clients to therapies.