Aims of classroom-centered research on second language learning and teaching are considered and contrasted with the experimental approach. Attention is briefly directed to methodological problems of experiments, such as controlling classroom events in various ways, and to conceptual weaknesses with study variables. In contrast, classroom-centered research, which involves observation, assumes that the description of classroom activities is a prerequisite to research on second language learning effectiveness and that formal instruction's contribution to language acquisition cannot be studied without measuring actual classroom activity. To illustrate the approaches employed by the experimental study and classroom-centered research, a hypothetical study about teacher correction of learners' errors is considered. Three vital roles played by nonexperimental investigations of the second language classroom are identified: (1) identifying variables whose importance for classroom learning should be investigated experimentally, (2) discerning the unique character of the second language classroom, and (3) developing new conceptualizations of the second language classroom. The view of the classroom largely guiding second language classroom research is that of an interaction between teachers and learners. Alternative views of the classroom have also arisen based on observation and analysis of the second language classroom. (SW)