In this paper on research into teacher behavior, the author considers first the current status of such research and the fact that the problem is so complex that no one knows or agrees what a competent teacher is. The difficulties of identifying and measuring teacher behaviors are discussed. The four broad areas of measurement (observation, student behavior and achievement, ratings based on recall, and personality tests) are considered in detail, together with additional problems including the noncumulative nature of earlier research; the problem of distinguishing, control and analysis of the variables; the rapid sequence of classroom events; the non-scientific nature of teaching; and teachers' lack of enthusiasm for researchers, particularly in inner-city situations. In an attempt to improve this situation, twenty-seven recommendations for future research are listed. The paper is based on Chapter 3 of the author's unpublished doctoral dissertation, Selected Teacher Behavior Attributes Rated as Desirable by Ninth-Grade Disadvantaged Students and Ninth-Grade Teachers of the Disadvantaged. (ED 043 563 and ED 043 564 are related documents. (MBM)