Battering has been described as an example of the total institution (Avni, 1991), within which batterers use violence, threats of violence, isolation, and other tactics of abuse to enforce compliance with their rules. Battering has well-documented physical, psychological, social, and economic consequences for women. It flourishes within the batterer-enforced privacy of the home. When it does come to the attention of outsiders, it is too often disregarded or trivialised - even when women seek protection through the justice system. In effect, this form of violence against women continues because it works. Drawing on both the literature and my own research, I argue that although there are various positive initiatives, single interventions are rarely enough to successfully breach the walls of the total institution and provide the resources needed for women to live independently of the men who batter them. Refuges (shelters) and advocacy programmes for women, stopping violence programmes for men, criminal justice policies mandating the arrest and prosecution of batterers, and the availability of civil-law remedies for battered women - none of these has proved to be the “magic bullet” (Buzawa & Buzawa, 1993b) which will, of itself, end battering. This is hardly surprising considering the resilience of the total institution and the multiple resources battered women require if they are to exercise a presumed choice to leave the batterer. I argue that a more comprehensive approach is needed in which multiple interventions are delivered in a consistent and co-ordinated manner with the twin objectives of (a) enhancing the safety and autonomy of women and (b) holding men accountable for their use of violence. Within the justice system, this can be achieved by legislative and administrative reforms which reduce the ability of decision makers to exercise discretion in woman-blaming and batterer-colluding ways, which ensure that there is a common set of priorities across agencies, which provide for the sharing of safety-relevant information between agencies and which include mechanisms for battered women’s advocates to monitor institutional practices so that decision makers can, in effect, be held accountable to battered women.