Qualitative research on batterer intervention programs (BIPs) has primarily consisted of interview-based studies of clients and facilitators. To date, no research has utilized observational data to understand how BIPs “work,” or the processes occurring in BIPs that promote prosocial behavioral change. Forty-four observations of BIP group sessions were conducted. Two key processes were found: “facilitator processes” (e.g., managing group dynamics and engaging clients in learning) and “client processes” (e.g., mutual aid, help-seeking, and support). More observational research on BIPs is needed to uncover the full range of processes occurring during BIPs and that can link group processes to client outcomes.