This longitudinal study examines to what extent different depressive and anxiety disorders and clinical characteristics are associated with subsequent weight change, while controlling for baseline weight, sociodemographics, health status, psychotropic medication use and (un)healthy lifestyle factors. Data are from a sample of 2447 respondents aged 18-65 years of the Netherlands Study of Depression and Anxiety (NESDA). Baseline depressive disorders and anxiety disorders were determined with the Composite International Diagnostic Interview (CIDI). Weight at baseline and after 2 years was measured and analyzed as continuous change score (mean change in weight 1. kg) and in categories of significant weight loss (<1. S.D. weight change equaling <4. kg), weight maintenance and weight gain (>1. S.D., >6. kg). After full adjustment for covariates baseline comorbid anxiety and depressive disorder and baseline Major Depressive Disorder (MDD) were associated with significant 2-year weight gain. Both current and remitted MDD at baseline and a baseline dysthymia, but none of the anxiety disorders, were associated with significant weight loss. This longitudinal study confirms a U-curved link between depression and weight change over 2 years. Furthermore, a dose-response effect of depression severity on 2-year weight gain was found.