OBJECTIVE:The study aimed to review the conclusion of a previously published meta-analysis which quantified distinct superiority of cognitive therapy to antidepressant drug-therapy (P < 0.0001). METHOD:We sought to include all studies used in the original meta-analysis. Adopting both that study's inclusion criteria and additional criteria resulted in a reduced set of studies. We analysed both 'completer' and 'intention to treat' data, using effect size and odds ratio quantification. RESULTS:There was an overall trend for cognitive therapy to be superior to antidepressant drug-therapy, but this was significant for only one of the four meta-analyses (an intention to treat analysis). We demonstrated considerable heterogeneity between studies, and a significantly higher drop-out rate in the antidepressant groups. CONCLUSION:The previous interpretation--cognitive therapy being distinctly superior to antidepressant medication--cannot be sustained from the currently analysed data set.