Objective: The purpose of this study was to: (a) investigate differences in demographic variables and psychological variables between treatment completers and dropouts among abusive women in a treatment program for domestic violence offenders; and, (b) create a predictive model that would correctly identify women at greatest risk of dropping out of the program. Method: The study employed a secondary analysis of 50 women, 25 treatment completers and 25 dropouts in treatment for domestic violence offenses. Results: Analysis indicated that very few of the demographic and psychological variables differentiated between treatment completers and dropouts. However, a logistic regression model was developed that correctly predicted treatment completion for 76% of the sample. Conclusion: Implications of the findings for improving retention rates among women attending batterer treatment programs were explored and discussed. (Contains 3 tables.)