Feedback by teacher comment, by television playback, and by self-analysis, singly, or together, reduced anxiety in subsequent performance as measured by nonfluencies in speech. Nonfluencies were counted in eight categories: the sounds, "ah,""um," or "uh;" correction; sentence incompletion; repetition; stutter; intruding incoherent sound; tongue slip; and omission. Performance in six treatment groups ranged from a low of 10.4 nonfluencies when there was videotape recording, teacher comment, and playback to a high of 37.1 nonfluencies when there was no videotape recording, no comment, and no playback. Since videotape playback without teacher comment resulted in almost double the nonfluencies (21.3) of those resulting from videotape playback with teacher comment, the conclusion was reached that a positive interaction between the two practices is to be desired. It is interesting that the mere presence of a television camera and recorder reduced student anxiety somewhat, perhaps, because of a "Hawthorne" effect. (MF)