Attachment theory, research, and assessments have become increasingly applied to settle child custody cases. We discuss such applications in relation to admissibility criteria for scientific evidence and testimony proposed by Faigman et al. (2014). We argue that attachment theory and research can provide valid "framework evidence"; group-based attachment research has yielded general principles suitable as a frame of reference for pertinent court decisions. In particular, child custody decision-making should generally be guided by research indicating that children benefit from attachment networks. In contrast, assessments of attachment quality fall short of providing valid "diagnostic evidence"; information that a specific individual/dyad is a "true" instance of a general group-level principle. In particular, such assessments do not yield valid information about whether a particular caregiver has better caregiving skills than another caregiver and will better support child development. We conclude that attachment theory and research should be admissible for framework but not for diagnostic testimony.