To assess the clinical utility of batterer typologies and to determine the extent to which mental health practitioners could successfully classify batterers into sub groups, data were obtained from 49 male batterers who were treated for domestic violence in an outpatient mental health facility. Scores on three validity scales and the 10 clinical scales of the Minnesota Multi-Phasic Personality Inventory (MMPI) were used to create empirically derived batterer subgroups. A preliminary comparison of the empirically derived typology with a theoretically based typology revealed many batterers were classified differently by the two strategies, although overall grouping strategies were similar in content. Further, mental health professionals had substantial difficulty choosing which of the cluster analytically created subgroups into which to sort MMPI profiles of individual batterers. Finally, different treatment efficacy indicators differentiated among the men in the two subtyping systems, suggesting each system offers unique clinical information. 20 references and 3 tables