An association between re-assault and mental health problems among batterer program participants has been increasingly documented by research and implied in batterer program guidelines. The majority of guidelines recommend assessment for such problems and referral to mental health treatment in addition to batterer program counseling. However, there is little documentation of assessment results and treatment outcomes. A research project was conducted to investigate the extent of mental health screening, referral compliance, and treatment effectiveness. The project included a formative evaluation of referral implementation, a service-delivery evaluation of the screening and referral, and an outcome evaluation of supplemental mental health treatment on batterer program completion and re-assault of the men's female partners. The formative evaluation exposed a few unexpected disruptive events and inconsistencies in referral procedures related in part to administrative turnover and differing priorities among agencies. The implementation modifications resulted in three stages of referral: voluntary referral, transitional referral, and mandatory referral.The service delivery evaluation showed nearly half of the batterer program participants (N=479 of 1043) screened positive on the Brief Symptom Inventory (BSI) and were referred to a local mental health clinic. A concurrent validity test revealed a correlation of the BSI with a more comprehensive screening instrument and a briefer one (n=93), but the test-retest reliability of the BSI was low with a decrease in positive screens over time (n=98). There was also little association with the BSI subscale results and the clinical evaluations obtained by referred men (n=38), and nearly 40% of the clinical diagnoses were for an adjustment disorder not warranting further treatment. Only 30% of the referred men received an evaluation, and 20% obtained some treatment under mandatory referral. Interestingly, at least a third of the referred men acknowledged a need for treatment and was more likely to obtain treatment.The outcome evaluation, based on a 12-month follow-up with female partners (65% response rate; n=308), produced no apparent effect of "intention to treat," represented by mandatory referral, on program completion or re-assault and other abuse indicators. (Re-arrests for violent and other types of crimes were substantially lower for the mandatory referrals in a confirmatory subsample of 300 subjects.) However, there was some preliminary evidence of a "dose response" for evaluated and treated men. Overall, the referral compliance was relatively low, but did increase under mandatory referral reinforced by a system-coordinator and case-manager. Sanctions for non-compliance remained inconsistent and may have affected referral compliance. The results reinforce recent studies exposing the challenges in establishing coordinated community response, and they question the utility of elaborate referral procedures for mental health referral. Alternatives might consider a mo...