Since its development, there has been growing utilization of the Safewards package of interventions to reduce conflict and containment in acute mental health wards. The current study used the opportunity of an implementation of Safewards across one large metropolitan local health district in New South Wales Australia to evaluate change. Specific aims of the study were to measure, for the first time in Australia, changes in shift-level reports of conflict and containment associated with Safewards introduction, and to measure any association with change in the violence prevention climate using a tool validated for use in the current study setting. Eight of eleven wards opted-in to participating in Safewards. Implementation was conducted over a period of 24 weeks (4-week preparation, 16-week implementation, and 4-week outcome phases). Conflict and containment were measured using the Patient–Staff Conflict Checklist Shift Report and violence prevention climate using the VPC-14. From 63.2% response rate, the mean (SD) reported conflict and containment incidents per shift fell from 3.96 (6.25) and 6.81 (5.78) to 2.94 (4.22) and 5.82 (4.62), respectively. Controlling for other variables, this represented reductions of 23.0 and 12.0%, respectively. Violence prevention climate ratings did not change. Safewards was associated with significant improvements in all incidents of conflict and containment, including the most severe and restrictive types, and this was largely unaffected by outcomes measure response rate, shift or weekday/weekend reporting, or number of ward beds. Safewards is increasingly justified as one of very few interventions of choice in adult, acute mental health services and should be widely utilized.