The diagnostic concept of Addiction to Near Death (ATND) is a construct with strong face validity. The purpose of this study was to empirically test the ATND construct in the context of a substance abusing population who were at high risk of contracting HIV. Interviews of female clients were conducted 30 days postdischarge from a residential therapeutic community. Clients were interviewed about their sexual activity and their condom use while they were in treatment. Results preliminarily support the ATND construct. The limitations were: data were from retrospective self-reports; the original study contained no baseline measures of sexual activity, safe-sex knowledge, condom use, HIV status; it had no male participants, no specific questions about near-death behavior, nor whether alternative safe-sex activities were practiced.