Background: Elderly people living in residential homes are at high risk for developing major depressive and anxiety disorders, and therefore deserve attention in terms of preventive interventions. We evaluated the feasibility and effectiveness of a guided self-help intervention for the prevention of depression and anxiety in these residents. Methods: We conducted a pragmatic randomized controlled trial in two parallel groups comparing the intervention with usual care in 14 residential homes in and surrounding the city of Amsterdam in the Netherlands. A total of 129 residents with a score of 8 or more on the Centre for Epidemiologic Studies Depression Scale (CES-D) screening instrument, who did not meet the full diagnostic criteria for disorders, and were not suffering from cognitive impairment were recruited between April 2007 and December 2008. Participants were randomized to a guided self-help intervention (n = 67) or to usual care (n = 62). The main outcome measures were improvement in the level of symptoms of depression and anxiety. The secondary outcome was improvement in participation in organized activities in the residential homes. The study is registered in de Dutch Cochrane Centre, under number ISRCTN27540731. Results: Only 21% of the participants (mean age 84.0 years (SD 6.7), 72.1% suffering from two or more chronic illnesses) completed the intervention. Although we found some large positive effect sizes on the CES-D, none of these effects was statistically significant. Conclusion: Although guided self-help may be promising in the prevention of depression and anxiety, it proved to be difficult to apply in this very old and vulnerable group of people living in residential homes. © International Psychogeriatric Association 2011.