Obstructive sleep apnoea (OSA) is the periodic reduction (hypopnoea) or cessation (apnoea) of breathing due to narrowing or occlusion of the upper airway during sleep. The mainstay of nocturnal treatment is continuous positive airway pressure (CPAP) therapy at home at night to prevent the resulting apnoea, hypoxia and sleep fragmentation which are believed to cause the symptoms of OSA. A number of clinical trials have evaluated the effectiveness of nasal CPAP) as a treatment. The objective was to review the effects of CPAP in the treatment of obstructive sleep apnoea in adults. We searched Medline 1966 to 1996, Embase 1974 to 1996, Cinahl 1982 to 1995 and the reference lists of articles. We consulted experts in the field. Randomised trials comparing nocturnal CPAP with placebo or other treatments in adults with obstructive sleep apnoea and an apnoea/hypopnoea index greater than five per hour. Trial quality was assessed and data were extracted independently by two reviewers. Study authors were contacted for missing information. Seven trials involving 177 people were included. All the studies had some methodological shortcomings. Six trials were of crossover design. Compared with placebo, CPAP showed significant improvements in several quality of life and depression measures. Patients preferred CPAP to placebo (odds ratio 0.4, 95% confidence interval 0.2 to 0.8). There was no significant effect on blood pressure or measures of sleepiness and sleep quality. Compared with oral appliances, CPAP significantly improved the apnoea/hypopnoea index (weighted mean difference -7.3, 95% confidence interval -10.0 to -4.7) and minimum oxygen saturation during sleep. Patients strongly preferred the oral appliance to CPAP (odds ratio 9.5, 95% confidence interval 4.3 to 21.1). CPAP appears to be more effective than placebo in improving some quality of life measures for people with obstructive sleep apnoea. It appears to be more effective than oral appliances in improving respiratory disturbances. People appear to prefer oral appliances over CPAP, but prefer CPAP over placebo.