The aim of the Follow-up Project Psychotherapeutic Communities is to describe the effects of psychotherapeutic treatment within a residential setting. The collection of data in this longitudinal study is mainly carried out by means of a mailed questionnaire. This implies that the researchers are faced with the problem of attrition. The decreasing respons-percentages stress the question of generalizability. In this article we report our investigation of the influence of attrition on the research findings of the project. The investigation leads to the conclusion that respondents and non-respondents do not differ on theoretically relevant variables such as well-being at the moment of discharge and readmission-rate. However, on a more specific level some minor differences are found which should be the subject of further investigation. It is stated that attrition does not challenge the generalizability of the results of the follow-up project.