Information used in this study was obtained from the standardized assessment of men referred to a male batterer treatment program. This assessment contained a 600-item history questionnaire that took approximately 90 minutes to complete. The information was then checked in an interview the following week, in which unusual responses were discussed, additional information was collected, and the Conflict Tactics Scale was administered orally. Of 598 men who completed the assessment, 526 were recommended for the treatment group. These 526 men were the focus of this study. The variables measures were in the following categories: demographics, family background and abuse history, history of antisocial and violent behavior, marital adjustment and relationship history, other adjustment problems, and court- mandated versus voluntary treatment. The study used a cross- validation design to protect against spurious findings. Of the 526 men recommended for treatment, only 41 percent attended a single treatment session, and only 25 percent completed the brief (10-week) treatment program. The variables associated with attrition fell into two general categories: those associated with lifestyle instability (e.g., moves, unemployment, youthfulness) and those variables that showed a congruence between the clients' self-identified problems and the targets of treatment (e.g., self-admitted problems with spousal assault). Suggestions are offered as to how programs could reduce their attrition rates by attending to the issues of client-treatment congruence and lifestyle instability. 1 table and 54 references