Several general theories of crime were tested with path models on 293 female prisoners in a U.S. State prison. The theories tested included Social Bond and Control, Thrill/Risk Seeking, and a new attachment-based Developmental Dynamic Systems model. A large battery of different instruments ranging from measures of risk taking, to a crime addiction scale, to Childhood Adverse Events, to attachments and clinical issues were used. The older general theories of crime did not hold up well under the rigor of path modeling. The new dynamic systems model was supported that incorporated adverse childhood events leading to (a) peer crime, (b) crime addiction, and (c) a measure derived from the Attachment and Clinical Issues Questionnaire (ACIQ) that takes individual differences in attachments and clinical issues into account. The results were discussed in terms of new approaches to Research Defined Criteria of Diagnosis (RDoC) and new approaches to intervention.