This paper is concerned with issues related to reasons why some early childhood intervention programs may not have lived up to expectations. By intervention is meant the introduction of planned programing deliberately timed and arranged in order to alter the anticipated or projected course of development. Although an intervention program should be a hypothesis testing program (one that requires that the dependent as well as the independent variable be clearly articulated) until the data is in, this is not the usual case. Matters recommended for discussion in establishing an intervention program include: (1) the variety of subsystems comprising the human organism, (2) the relationship between systems longitudinally, (3) the when of intervention, i.e., identification of the significant epochs wherein particular interventions would have more relevance in getting the child on an appropriate "course" of growth, (4) the degree to which the child can assimilate and accomodate to the program, (5) the degree to which certain competencies or structures are susceptible to modification, (6) the relationship between the child and the broader environmental context, and (7) the interaction between the child and his parents. It is concluded that an airing of these matters will not only make intervention experience of value to the target children (minority group members), but will also facilitate the child's subsequent learning. (CK)