Background: Computer-guided therapy is an innovative treatment strategy that could have an important role in the future of psychological treatment. This paper summarises the available published evidence that assesses the effectiveness of a computerised cognitive behaviour therapy (CCBT) for obsessive-compulsive disorder (OCD). Methods: Fifteen electronic bibliographic databases including Medline, Embase, the Cochrane Library, Cinahl, PsycINFO, Biological Abstracts, HMIC and NHS CRD databases were comprehensively searched in March 2004: [‘obsessive compulsive disorder’ (text and indexed terms)] AND [‘cognitive therapy’ (text and indexed terms)] AND [‘computer’ (text and indexed terms)]. Reference lists of included studies, guidelines, generic research, trials registers and specialist mental health sites were hand-searched. Results: The search produced 149 citations from which we identified two RCTs and two single-arm studies with relevant data. All four studies used one software programme – BTSteps.In the large RCT, YBOCS effect sizes for BTSteps, therapist-led cognitive behaviour therapy (TCBT) and relaxation (RLX) were 0.84, 1.22, and 0.35, respectively. The smaller RCT found significantly better outcomes with brief scheduled support compared to brief on-demand phone support. Conclusions: BTSteps was as good as TCBT for reducing time spent in rituals and obsessions and for improving the Work and Social Adjustment Scale (WSA), and was superior to RLX treatment. The available evidence also showed that improvement of OCD persisted beyond the end of CCBT. TCBT was more effective than CCBT for all patients overall though not in those who went on to start self-exposure. Such a system has the potential to widen the access to CBT in general and considerably shorten clinician-guided care.