Potential respondents included all men who had attended at least one counseling session at the House of Ruth in Baltimore, Md., and were not currently attending the program. Of the 312 men presumed to have received questionnaires, 53 returned usable ones. The self-report questionnaire was designed to elicit information on violent behaviors exhibited by the batterer toward a female partner. The inclusion of the Conflict Tactics Scale in the questionnaire contributed to this purpose. The counseling involved various activities intended to force a critical examination of the batterer's attitudes and to help batterers develop constructive ways to communicate their feelings other than through violence. Although the respondents were demographically heterogeneous, the majority had entered counseling voluntarily, were married at intake, were living with their partners at termination of counseling, and were Protestants. The recidivism rate for the sample was 35 percent. Recidivism was not significantly influenced by the frequency of attendance at counseling, whether or not attendance was voluntary, or drug use. Those most likely to recidivate were living with their partners at the termination of counseling, had problems with alcohol, and had witnessed violence between their parents. The small sample size, the limited credibility of batterers' self-reports, and the low completion rate makes these findings provisional. 2 tables, 16 footnotes.