To determine which types and qualities of science teaching strategies stimulate students to become involved in learning, middle and high school students (N=147) were observed for time-on-task (time actively spent by the student in learning) and off-task during periods while being taught by five student teachers. The "Teaching Strategies Observation Differential" (TSOD), rating teacher style on a scale from 0-10 (extremes defined as expository/direct and inductive/indirect), was used to measure verbal and non-verbal classroom activities. The quality of implementing strategies (lecture, direction, questioning, demonstration, investigation, response, or teacher guidance) identified by the TSOD was measured by an instrument developed for this study (included in an appendix). In addition, the "Test of Logical Thinking" (TOLT) was administered to determine possible relationships between developmental level and time-on-task. Results indicated that not only the type of teaching strategy, but also the quality of implementing that strategy have an influence on student time-on-task, suggesting that teacher education programs emphasize both a repertoire of strategies and the quality of implementing each type. Results also indicated that there was a relationship between formal reasoning ability and time-on-task, suggesting that teacher education programs also emphasize cognitive development theory and applications of this theory to instruction to facilitate learning. (Author/JN)