One Teaching Laboratory (TL) task, a component concerned with several behaviors related to classroom questioning, was developed and tested. Subjects were 86 teacher candidates enrolled in four sections of an introductory teaching course. Two experienced instructors each taught two sections, one section utilizing the TL component and the other not. Identical numbers of TL and non-TL Ss (N-43) participated in four 2-hour class sessions. The TL treatment, using specifically prepared TL manual, evaluation forms, and listening guide), consisted of presentation of questioning strategy concepts and participation in a microteaching cycle: Ss taught a lesson, evaluated and discussed it, and retaught it. The non-TL treatment consisted of presentation of the questioning strategy concepts and of Bloom's "Taxonomy"; an educational game called "Questionize"; and teaching a microteaching lesson. Audio tapes of microteaching lessons of all Ss on the final day constituted the data source. The 24-category Questioning Strategies Observation System (QSOS) was used to code and analyze tapes. Multiple linear regression procedures were employed. Findings indicate that the TL treatment was more effective than the non-TL in influencing candidates (1) to ask more questions of a cognitive rather than affective or procedural nature and (2) to react to pupils' responses to questions in a more positive (accepting, supporting) manner. (Implications are discussed.) (JS)