The stated purpose of this paper is to review some of the critical issues that arise in the planning and conduct of the evaluation of action programmes amongst the socially and educationally disadvantaged, and to suggest the means that might be adopted to overcome the inherent difficulties. The discussion first focuses on the characteristics of social and educational action programmes. Then, the functions of evaluative research are discussed. Following this, the discussion asserts that the coexistence of educational and social disadvantage in extreme forms tends to evoke programs which make either one or several interventions and assume that the interventions will make the social system work more effectively. It is advocated here, that in the planning stage, a model be constructed into which the proposals for change can be fitted and, through which, how they work and their intended effects can be seen. The value of such models, it is stated, is that they allow each proposal to be worked out in terms of its specific objectives, the agencies of the system and the instruments of action. The discussion then focuses on "the choice and formulation of objectives" and "the special status of evolving programmes". Finally, the discussion highlights considerations regarding "the design of evaluation experiments" and "the relationship between research and action".