Reported is a comparison of the effectiveness of two instructional methods of preservice elementary teachers in teaching science as well as a comparison of the relationships of initial attitudes toward teaching science and science process skill competencies to initial teaching behaviors. Forty lower elementary preservice teachers were randomly assigned to either an instructional method designed to improve the skills of observing, classifying, and inferring, or to an instructional method designed to improve directly the teachers' ability to elicit these behaviors in children by questioning. Measures of pre- and post-instructional attitude and of behavior during science teaching were made. Results indicated that even prior to instructional treatment, teachers performed a range of desirable behaviors including a high question complexity level, and that knowledge of the processes of science had precedence over attitude toward teaching of science. There were no significant differences in the effects of the two instructional treatments. The Process Questioning Strategies instructional method was effective in reducing the amount of teacher controlled silence including use of demonstrations. The Process Skill instructional method by itself was effective in increasing the proportion of time teachers devoted to questioning and allowing student response. (EB)