This paper examines current forms and instruments of classroom observation and suggests directions for future research studies. The following topics are covered: the need for good criterion measures, the use of observational instruments in outcome measures, the development of procedures to code the content that is covered in a classroom, the dangers of excessive complexity, the use of observation for teacher competency assessment, naturalistic observation, the coding of questions and cognitive interactions, a typology of questions and cognitive interactions, the indexing of implementation, and a proposal for a data bank for secondary analyses. (JA)