This study qualitatively examined the perspectives of clinical social workers on non-offending mothers of sexually abused children. The study examined whether clinicians still used collusion to explain mothers' behavior, despite research refuting collusion. Findings revealed that, although workers did not use collusion, they still constructed mothers negatively. Multiple contexts of agency practice influenced constructions. Administrative use of authority to implement external constraints led to workers' resistance, which involved humor with gender and ethnic components. The agency's role as a graduate social work teaching site contributed the following: Field instructors transmitted the belief that incest typified severe family difficulties and posed complex assessment and intervention problems. Implications for effective practice are discussed.