Alternatives to criminalization for the simple possession of illicit drugs are increasingly of interest to policy makers. But there is no existing theoretically based, empirically tested framework that can inform development and evaluation. This article presents a realist programme theory of such alternatives. It bases this on a realist review, which followed the Realist and Meta-narrative Evidence Syntheses: Evolving Standards (RAMESES). It describes the systematic process of searching the literature in English on nine relevant countries (Australia, Czech Republic, Denmark, Germany, Jamaica, Netherland, Portugal, the UK, the USA) for information on alternative measures in three categories: depenalization; diversion; and decriminalization. It shows how these measures – in theory and in practice – combine with pre-existing social conditions and institutional contexts to trigger mechanisms across three causal pathways (normative; criminal justice; and health and social services). It shows how some posited causal processes are more empirically supported than others. Alternative measures can reduce harms imposed by criminal justice processes without increasing drug use or related health and crime harms, but this depends on specific combinations of contexts, mechanisms and outcomes.