Teacher-child relationships are complex phenomena which can be described and observed from a variety of angles, on many different levels. Recent research contributes some information applicable to the analysis of teacher-child relations and identifies issues for further study. Teacher functions or duties may depend, in part, upon the teacher's background and the demands of her program. Although four types of functions are defined (maternal, therapeutic, facilitator, instructional), no research has yet been done to investigate teachers' time distribution among these role functions. While the characteristics of teachers can be examined in two broad classes, attributes and behaviors, the research is difficult to synthesize and summarize. However, it provides indications that teachers may need help in developing skills to extend the information processing abilities of pupils and to build more constructive classroom climates. These skills need to be clearly identified. Research involving larger sample sizes is needed to support studies assessing the predictability of teacher behavior from the specifications of curriculum models. Few studies deal with the effects of teachers on children. Welcome additions to the existing body of concepts are expected from the Planned Variation Experiment with Head Start curriculum. References are given. (WY)