Some of the evidence presented by Laws against CBT for schizophrenia appears to be more sobering than the evidence that exists for CBT for CFS, but there are interesting implications of the types of evidence he considers. For example, he goes into the effects of researcher allegiance and (the lack of) blinding, which remove the small or marginal effects that the evidence of CBT apparently had for schizophrenia. These issues have not been explored in CBT for CFS, where most studies are done by staunch proponents of the therapies and almost all are non-blinded. Re "Perhaps the most remarkable fact is that over 75% of published studies document no significant reductions in positive symptoms, negative symptoms or in the key symptom of hallucinations." In contrast to the above, most CBT trials for CFS do show statistically significant improvements in self-reported fatigue. Not sure the same applies to self-reported physical functioning, several do but the last Cochrane review (which needs updating) did not show any obvious effect, and neither did the latest NHS data on CBT for CFS. The PACE Trial did, but also showed that the effects on pain and other self-reported symptoms are minimal at best. With the more objective measures, the overall evidence suggests no significant improvements in physical activity levels, 6-minute walking test distances, employment hours, welfare dependence, insurance payouts, or total service use. CBT (and GET) are based on the controversial hypothesis that the symptoms and disability of CFS, regardless of initial triggers, are primarily perpetuated by maladaptive beliefs about the illness, fear avoidance of activities, hypervigilance to normal bodily sensations, and the physiological consequences of deconditioning and circadian rhythm disturbance, etc. CBT for CFS appears to modestly help a minority, so how has it been oversold? 1) Provides patients with an unproven "explanation" for their symptoms, which fails to account for the abnormal biological responses to exercise which cannot be explained by deconditioning and are not generally found in deconditioned controls. 2) It is commonly claimed by proponents to substantially benefit the majority of patients e.g. 70%, but the net clinical response rates (experiment rate minus control rate) are more like 15% (Cochrane, PACE Trial) so a NNT of about 7 for a modest improvement. The 'recovery' rates from the PACE Trial are misleadingly exaggerated, since the thresholds for normal fatigue and normal physical function overlapped with trial criteria for severe chronic disabling fatigue. On closer inspection, the additional criteria for recovery may indicate an improvement but certainly do not guarantee a 'recovery' in the normal sense of the word. Erroneous hype about 'recovery' in the Lancet was ruled misleading by the Press Complaints Commission. 3) It is usually assumed that because CBT/GET aims to increase activity levels and function, then it must occur during therapy. But the more objective measures do not support these assumed improvements. There is no good evidence of genuine recoveries that have been road tested with the normal activities of daily life. 4) Similarly, there have been a lack of consideration for whether the effects are genuine improvements in health or the result of the placebo response and changes to 'questionnaire-answering behaviour' that may arise due to encouraging patients to reinterpret symptoms, believe their illness is not as serious as they thought, and that they should focus less on symptoms and limitations. The Cochrane 2000 review suggests that CBT is not superior to oral placebo in a trial which also tested biomedical treatment, and the Cochrane 2008 review suggests that CBT is not superior to waiting list control. 5) It is supposed to be safe, but patient surveys suggest otherwise, although these results are dismissed as being the selective informal results of improperly applied CBT/GET in the real world. For about 20 years this safety was presumed in the literature based on group average scores and not from recording adverse effects on an individual level. The PACE Trial was supposed to resolve this concern, but the definition of adverse effects were generally much stricter than for the post-hoc definitions of improvement, and there is unpublished data. The issue of safety is also tied up with that of objective outcome measures: CBT/GET appears to be generally 'safe' as long as patients are not pushed beyond their limits, but contrary to what we are told by proponents of the therapies, the evidence suggests that CBT/GET do not overcome these limits. 6) People who dispute the rationale and claims of effectiveness of CBT (and GET), or challenge questionable post-hoc protocol changes in 'definitive' trials e.g. PACE, are often dismissed as anti-psychiatry extremists who don't understand the mind-body connection, fear the stigma of mental illness, ideologically resist suggestions that CFS is not biomedical, etc. This dismissal has consequences for scientific understanding and for patients who fail to improve. In the FINE Trial which tested similar therapies based on a similar rationale, patients were often regarded by the nurses as the bastards who didn't want to get better, which is unsurprising given the non-representative treatment model which they were trained to deliver.