For the public and for most policy makers there is no doubt that the career of heroin addiction necessarily involves gradually losing everything one has: health, income, a home, a stable network of relatives and friends, and, last but not least, any sense of self as a moral agent. Although this view is factually incorrect insofar as there are examples of controlled heroin use without such negative effects (cf. Zinberg, 1984), the image of the addiction career as a necessarily downward spiral is represented in the literature as well. The clearest example is the medical model of addiction, in which the addiction career is described as a process of inevitable social deterioration: the drug addict becomes more and more involved in deviant and illegal behavior. The addiction career can be terminated only by therapy or death. Other literature, however, describes the course of addiction as a process of maturing out (Prins, 1995), in which a considerable number of addicts are able to kick the habit by themselves, step by step (Winick, 1962). The latter implies that addicts are able to control their career to a certain extent. The two approaches can be represented by different metaphors: downward spiral and controlled descent. This article, which is based on 48 interviews, discusses how heroin addicts portray their careers and the extent to which they control the downward process. It is supposed that information of this kind will enhance our knowledge of addicts' methods of controlling their situation, which can be relevant for policy and therapy. Our main question, therefore, is whether addicts portray themselves as victims of forces beyond their control or, at least partially, as capable of regulating their situation. Methods In the frame of a more extensive study of differences between heroin addicts in and outside treatment settings in the city of The Hague (The Netherlands), interviews were conducted with 48 addicts who had been using heroin regularly (three days a week or more) for at least two years. Among the respondents were eight women. The sample consisted of 21 heroin addicts who had not been in contact with drug treatment agencies for at least two years, five clients of an outpatient methadone program, 11 participants in a clinical detoxification center, and 11 addicts staying in a drug-free therapeutic community. The 21 non-treatment respondents were selected by a snowball sampling method (see ElandGoossensen et al., in press). The treatment group were interviewed between two and four weeks after admission. Respondents were invited to tell about how heroin use was embedded in their lives. Experiences with treatment and, if relevant, reasons for rejecting and quitting treatment were also explored. In the interviews we used a topic list corresponding to the life areas that are distinguished in the Addiction Severity Index (McLellan et al., 1981; Hendriks et al., 1993); these include chemical abuse, medical, psychological, legal, family/social, employment/support. Housing was added as a topic. Obviously, in the respondents' stories the different life areas were not so clearly separated. Interviews took approximately one hour each and were audio-taped. The interviews were transcribed. Results are presented in three sections. First we discuss to what extent respondents portray their careers as dominated by forces beyond their control. Next we present data on respondents' descriptions of controlling behaviors. Finally we show how personal values are related to control and, consequently, are of influence on the addiction career. Downward spiral In response to the request to tell about how heroin use was embedded in their lives, all respondents spontaneously described aspects of their careers as (part of) a downward spiral. The following extract from the interview with a respondent in the detox group is an example: I: Our first question is . . . What has been the meaning of heroin use in your life, what has been its influence? …