The study of teacher effectiveness is confronted by a number of problems that are generally associated with the conduct of behavioral research. It is possible in some instances to resolve or circumvent some of the current methodological stumbling blocks that tend to reduce the credibility of research findings. This paper discusses three methodological problems: (1) the importance of the teacher relative to his ability to affect student growth; (2) the attempts to operationalize constructs that appear to be related to student outcomes; and (3) the statistical problems associated with measuring student growth. Several alternative solutions to these problem areas are presented. (JMF)