Problem gambling is of serious public, social and clinical concern, especially so because ease of access to different types of gambling is increasing. A systematic review and meta-analysis was carried out to determine whether Cognitive-Behavioural Therapies (CBT) were effective in reducing gambling behaviour. Twenty-five studies which met the inclusion criteria were identified. Overall, there was a highly significant effect of CBT in reducing gambling behaviours within the first three months of therapy cessation regardless of the type of gambling behaviour practiced. Effect sizes were also significant at six, twelve and twenty-four month follow-up periods. Sub-group analysis suggested that both individual and group therapies were equally as effective in the 3 month time window, however this equivalence was not clear at follow-up. All variants of CBT (cognitive therapy, motivational interviewing and imaginal desensitization) were significant, although there was tentative evidence that when different types of therapy were compared cognitive therapy had an added advantage. Meta-regression analyses showed that the quality of the studies influenced the effect sizes, with those of poorer quality having greater effect sizes. These results give an optimistic message that CBT, in various forms, is effective in reducing gambling behaviours. However, caution is warranted because of the heterogeneity of the studies. Evaluation of treatment for problem gambling lags behind other fields and this needs to be redressed in the future.