Abstract
In the treatment of dnig abusers a continuing problem is the high attrition rate. This study focussed on m New South Wales treatment facility, Odyssey House, and aimed to discover some of the variables associated with attrition. A simple of people admitted to the program was divided into two groups, those who left within 30 days and those who did not. Most of the client variables were nonsignificant. The most significant predictor of attrition was the client's status IS a new admission or as a readmission; re-admissions were more likely to stay than newly admitted people. Also significant were the length of previous treatment, suicide attempt history, clients' perceptions of the toughness of the program and the predictions made by the admission team. The study indicated a need for further investigation of program and environmental variables and of interactions among these and client variables.