The recent redevelopment of the alcohol and drug service system in Victoria, Australia, included the creation of a new residential service type, alcohol and drug supported accommodation (A&DSA) that is based on the principles of halfway houses. In 1999/2000 a process evaluation of 19 services was conducted. A mixed methods approach involved: a service questionnaire; interviews/focus groups with staff, clients, and significant others; site visits; a file audit; service monitoring data, and interviews with key stakeholders. The evaluation showed that service clients are polydrug users, on welfare benefits, and homeless or in insecure housing before entering A&DSA. Fundamental service’ elements include secure accommodation, individual plans, limited support, and service linkages. Most services met or partially reached set targets regarding client numbers, support plans and goal achievement. Services with a target average length of stay beyond 3 months found their target difficult to achieve. A substantial minority of services did not meet the required equal representation of males and females. Enriched models have been developed for special needs groups, including women, parents with dependents, young people, the homeless, and significant others. The evaluation highlighted the adaptability of the A&DSA model and the need for policy responses to address current resource constraints.