The purpose of this study was to compare the effects of two prereading instructional treatments on students' comprehension of narrative and expository texts. The authors randomly assigned 63 fifth-grade students in the U. S. to three groups. Each group was exposed to three treatment conditions. The three conditions were (a) a teacher-directed condition in which teachers read prepared scripts designed to provide students with important information necessary for understanding upcoming texts, (b) an interactive condition in which teachers activated and discussed students' prior knowledge about the topics of upcoming texts, and (c) a control condition in which no prereading instruction was provided. Three narrative and three expository passages were counterbalanced with the three treatments and three teachers across groups, to control for effects of instructor and passage. Hierarchical multiple-regression analyses of within- and between-subject effects indicated that the teacher-directed condition was more effective than the interactive condition at promoting comprehension, and that both treatment conditions were superior to no prereading instruction at all. The teacher-directed condition may have been more effective because (a) it focused only on the most important information necessary for understanding the text, (b) it included direct and explicit instruction, or (c) it was more typical of traditional prereading instruction and therefore more familiar to students. However, the interactive strategy may be more useful for other instructional goals, such as developing students' abilities to activate their own background knowledge.