This study examines the effectiveness of a group CBT (CBGT) intervention in reducing a variety of symptoms and problem areas associated with social anxiety disorder. A longitudinal cohort design assessed changes in standardized psychological scales assessing general mood and specific aspects of social anxiety. Questionnaires were completed pre-programme (time 1, n = 252), post-programme (time 2, n = 202), and at 12 months follow-up (time 3, n = 93). A consistent significant pattern was found for all variables: pre-intervention scores were significantly higher than both post-intervention scores and 12-month follow-up scores. Large effect sizes were found and rates of clinical significant changes varied, with over half of the participants recording clinically significant changes in general mood. Individual CBT can be translated successfully into a group format for social anxiety. Given the high completion rate, the intervention is acceptable to participants, feasible, and effective in a routine clinical service.