There are two significant difficulties in building a general criminal justice theory. First, different from criminology theories, criminal justice produces multiple outcomes at different levels. Second, the scopes of existing theories largely originate from Western contexts and data, few including cross-cultural variation. This paper outlines a unified theory to explain multiple criminal justice outcomes at the system, institutional, and individual levels across cultures under a paradigm shift from the current “monotonic paradigm” to a more general “comparison paradigm.” The new paradigm logically contains the existing paradigm while broadening research questions and scope of criminal justice studies. It constructs a new set of concepts and propositions, presenting an effort toward a general causal criminal justice theory.