It is now common for courts to mandate partner-assaultive men to attend batterer intervention programs. While some data point to the effectiveness of such programs, other data indicate that between 41 to 90 percent of men initially referred for treatment either do not attend or prematurely drop out. It is important to understand the characteristics of men that do not complete court-mandated batterer intervention programs (BIP's). This study investigated 199 men ordered by the Dallas County Domestic Violence Court to attend BIP. They were administered a computer-assisted structured interview assessing stages of change, processes of change, psychological distress, relationship conflict, and other characteristics prior to BIP. Also, 60 female partners of male participants reported on new instances of physical and emotional abuse, perceptions of safety, risk for future violence, and perceptions of male change. Five assessments were conducted over a 13-month period. The results indicated that 40 percent of men mandated to attend BIP did not complete their program, 27 percent of men were rearrested, and 62 percent reported new acts of male-to-female violence. Cluster analysis revealed four to five distinct stages of change groupings, with all but one of those clusters representative of men that were either mildly or reluctantly predisposed toward change. Men of African-American race, and higher recontemplation scores predicted BIP attrition. High precontemplation, low action, more alcohol problems, and higher levels of anger prior to treatment predicted rearrest. Men in the Borderline/Dysphoric and Generally Violent/Antisocial subtypes were more likely to drop out of treatment and be rearrested. 18 tables, 10 figures, 78 references