The requirements for a rigorous cost‐effectiveness analysis are considered. The evidence base on services for people with ID is then reviewed with an emphasis on the transition from institutional services to those in the community and postdeinstitutionalization research on the costs and outcomes of different forms of community provision. Despite certain limitations, a number of conclusions are possible: (a) overall, community services provide for a better quality of life than institutional care; (b) whether community service costs are more or less than institutional costs may depend on factors such as funding mechanisms, wage rates and the level of investment in institutional services prior to deinstitutionalization, that are likely to differ between countries; (c) economies of scale are not pervasive but may arise among settings with very small group sizes depending on staffing model; (d) the costs of staffing are a major element in total service package costs and there is scope to relate staff input more precisely to the needs and characteristics of service users; (e) the factors responsible for variation in costs and outcomes are incompletely understood; (f) however, resource input does not appear to result in enhanced outcome; (g) greater staff input is inefficiently translated into performance which affects service users; and therefore, (h) operational culture and staff training and management are important determinants of ultimate outcome. More research is required on the factors that drive costs on the one hand and outcomes on the other. An international consensus is required on the important variables to be described when doing research on residential support arrangements, so that relationships between environmental characteristics and either costs or outcome can be identified with greater confidence.