A study designed to determine what characterizes effective instructional explanation in reading involved reading instruction, by the researcher, of four reading groups and one child in a third/fourth grade classroom in the fall, and two groups plus one child the remainder of the academic year. The researcher was observed 32 times during the year; field notes were taken and lessons were audio tape recorded and/or video recorded and then transcribed. Answers were sought for the following research questions: (1) What characterizes the research participant's explanations during reading instruction? (2) What characterizes the teacher-student interactions during the research participant's instruction? (3) How were the lessons organized? (4) Is there any evidence regarding the effectiveness of explanation in creating student outcomes? Results of the study have provided rich descriptive data to conceptualize the nature of instructional explanation, including three aspects of instructional explanation that go beyond a simple concept of expository teacher talk. Data revealed no relationship between instruction and student gains, and a disparity between what the research participant did when explaining as a practicing teacher and the way he conceptualized explanation behavior for his methods-course students. The findings suggest that instruction is much more complex than models, opinion, or much of the previous research have indicated, and that theoretical models of instruction have little or no reliability as far as identification of essential and critical variables. (EL)