This study developed and evaluated a manual for the improvement of teachers' effectiveness in explaining. A class of 60 beginning teacher trainees was randomly divided into two groups. All subjects were given a pretest that required them to make three oral explanations in response to three questions asked by an experimenter; the questions and answers were tape-recorded. Following the pretest, half of the class received the "How To Explain" manual as the basis for training; the other half did not. After training, all subjects were given a posttest consisting of three new questions. Randomly ordered answers for each subject were played to 10 raters, who were seventh- and eighth-grade students. The students rated each answer on four dimensions: a) organization, b) clarity, c) quality, and d) rank among the six answers for that subject. Subsequently, six experimental and six control subjects were randomly selected, their posttest answers only were transcribed, and these answers were rated and coded by trained judges on six dimensions a) content validity, b) logical organization, c) verbal emphasis, d) rule-example-rule, e) lesson divisions, and f) vagueness--in an attempt to determine whether the explanations of the experimental group reflect the emphasis on these six dimensions in the "How To Explain" manual. Results indicate that the manual and the total program in which it was used had significant effects in the experimental group's explanations. A 35-item bibliography and appendixes are included. (Author/MJM)