Outcomes for cognitive-behaviour therapy (CBT) in randomised controlled trials (RCTs) have rarely been compared to those in routine clinical practice. Taking the case of CBT for chronic fatigue syndrome (CFS), we evaluated the results of a successful RCT against those of the same treatment given in the same setting as part of routine practice. Fatigue and social adjustment scores were compared for patients who received CBT for CFS as part of a RCT (N=30) and patients who received CBT as part of everyday clinical practice (N=384). The results in the RCT were superior to those in routine clinical practice. Between pre-treatment and 6-month follow-up, the RCT showed a larger reduction in fatigue and greater improvement in social adjustment than those in routine treatment. The changes in fatigue scores were similar for both groups during treatment but were greater in the RCT between post-treatment and follow-up. Potential reasons for the superior results of the RCT include patient selection, therapist factors and the use of a manualised treatment protocol. Practitioners need to pay particular attention to relapse prevention and ensuring adequate follow-up in addition to encouraging patients to continue with cognitive-behavioural strategies once treatment has ended.