The transtheoretical model of change addresses men's change in attitudes by classifying them into stages based on their change motivation and behavior. Men with no motivation to change and who are not engaging in any attempts to change are classified as in the precontemplation stage. Because these men do not believe that they have a problem that needs to be changed, they are at high risk of dropping treatment at high rates. The next stage, contemplation, refers to men who acknowledge that they have a problem and are working toward developing sufficient change motivation to modify contributing behaviors. These men are still somewhat likely to drop treatment, particularly if actual changes in behavior are expected or required. Men in the final stage of change, the action stage, are both motivated and actively involved in making attempts to change their problematic behavior. These men are most likely to be "ready" for intervention and are therefore expected to have the lowest rates of dropout. In order to test this theory with men who have physically abused their intimate partners, testing was done with men who attended the first session of treatment at a batterer treatment agency in southwestern Ontario, Canada, between March 1998 and May 1999 (n=345). A total of 308 men were included in the study sample. The treatment program was similar to the majority of batterer programs in its structure and counseling approach. Stage of change classifications were determined by administering the University of Rhode Island Change Assessment Scale. Counselors also provided categorical judgments about men's stage of change following a 2-3 hour intake interview. Of the 308 men who began treatment, only 119 (38.6 percent) completed the 17-week program. The remaining 189 men dropped out. The study found that counselor-rated, but not self-rated stage of change significantly predicted treatment completion once traditionally used demographic, contextual, and personality variables were taken into account. Men identified by counselors as being in the precontemplation stage were 2.3 times as likely as men rated in the contemplation stage and 8.8 times as likely as men rated in the action stage to leave treatment. Referral source, age, and history of arrest also were significantly related to the prediction of attrition. The best combination of predictors led to the successful classification of 72 percent of the cases (98 percent of the dropouts but only 19 percent of the completers). This report concludes with a discussion of the limitations of the study as well as the implications of the findings for the use of the transtheoretical model to predict attrition from batterer treatment programs. 3 tables, 50 references, and appended counselor rating instrument