We appreciate the thoughtful comments of Koesters et al. (1). We were of course aware of their meta-analysis, which focused on a subset of the published studies comparing venlafaxine and various selective serotonin reuptake inhibitors (SSRIs) and regret that their report was not published in time to be available for discussion in our article. With respect to the first point raised by Koesters et al., we agree that the differences between the results of their meta-analysis and ours are small. We also agree that publication bias can distort the results of meta-analyses; this is why we included all randomized double blind trials comparing venlafaxine and SSRIs in major depressive disorder conducted by Wyeth before 2007, whether published or not.