A series of primarily laboratory-based studies found attention bias modification in socially anxious participants to lead to reduced anxiety. It is argued that the failure to replicate the positive results of attention bias modification in the study of Carlbring et al. may be due to reasons other than the application through the Internet. A number of controlled studies failed to replicate the positive effects of attention bias modification in clinically rather than subclinically socially anxious subjects. Given the lack of robust evidence for attention bias modification in clinically socially anxious individuals, the author is inclined to consider attention bias modification as 'the Emperor's new suit'. Results achieved with regular Internet-based treatments for social anxiety disorder based on cognitive therapy and exposure methods are much better than those achieved with attention bias modification procedures delivered 'face to face' in clinically distressed participants. Given the lack of robust evidence for attention bias modification in clinical samples, there is no need yet to investigate the implementation of attention bias modification through the Internet.Please see related article: http://www.biomedcentral.com/1471-244X/12/66.